J. Steraid Biochem. Molec. Biol. Vol. 41, No. 1, pp. 11-20, 1992
0960-0760/92 $5.00 + 0.00 Copyright © 1992 Pergamon Press pie
Printed in Great Britain. All rights reserved
MECHANISMS OF INSULIN INHIBITION OF ACTH-STIMULATED STEROID SECRETION BY CULTURED BOVINE ADRENOCORTICAL CELLS NANCY A. KLEIN, 1. RICHARD N. ANDERSEN,1 PETER R. CASSON,1 JOHN E. BUSTERl
and ROBERTE. KRAMER2t Departments of ~Obstetrics and Gynecology and 2Pharmacology, University of Tennessee, Memphis, TN 38163, U.S.A. (Received 14 March 1991)
Summary--Results of previous studies indicated that insulin at levels comparable to those in humans during hyperinsulinemia decreased ACTH-stimulated cortisol and androstenedione secretion by bovine adrenal fasciculata-reticularis cells in primary culture. In the present studies this inhibitory action was examined further by comparing the effects of insulin on ACTH-stimulated corticosteroid secretion with its effects on 8-(4-chlorophenylthio)-cAMP (cpt-cAMP), forskolin- and [Sval]angiotensin II (Ang II)-stimulated corticosteroid secretion. Effects on corticosteroid secretion were correlated with effects on cAMP accumulation and rates of cAMP production. Monolayers were incubated for 24 h in the absence or presence of each agonist alone or in combination with insulin. Insulin (1.7 x 10-9 or 17.5 x 10 -9 M) caused about a 50% decrease in cortisol and androstenedione secretion in response to ACTH (10-'1 or 10-s M). Insulin also decreased ACTH-stimulated aldosterone secretion by cultured glomerulosa cells. Cpt-cAMP (10 -4 or 10 -3 M)-stimulated increases in cortisol and androstenedione secretion were inhibited by insulin, but to a lesser extent than those in response to ACTH. The inhibition of cpt-cAMP-stimulated steroid secretion was not related to increased degradation of the cyclic nucleotide. Increases in cortisol and androstenedione secretion caused by a submaximal concentration (10-6 M) of forskolin were decreased 50-70% by insulin. In contrast, insulin failed to significantly affect cortisol or androstenedione secretion caused by a maximal concentration (10 -5 M) of forskolin. The secretory responses to Ang II (10 -s M) were also unaffected by insulin. The effect of insulin to inhibit ACTH-stimulated steroid secretion was accompanied by a reduction in cAMP accumulation as well as an apparent inhibition of adenylate cyclase activation. These data indicate that the effect of insulin to attenuate ACTH-stimulated corticosteroid secretion results from both an inhibition of ACTH-stimulated adenylate cyclase activity and an antagonism of the intracellular actions of cAMP.
INTRODUCTION
Studies by Farah et al. [6] have extended these observations and demonstrated bimodal effects of insulin on D H E A production in vivo, Farah et al's findings are in agreement with the results of recent studies [7] from this laboratory in which ACTH-stimulated androstenedione secretion by primary cultures of bovine fasciculata-reticularis cells was stimulated by insulin at physiological levels (~< 1.7 pM) and inhibited by insulin at hyperinsulinemic levels (1> 1.7 nM). High concentrations of insulin also inhibited ACTH-stimulated cortisol secretion. The present study was undertaken to identify the mechanism(s) underlying the inhibitory action of insulin on ACTH-stimulated androstenedione and cortisol secretion. It is generally accepted that the steroidogenic effect of A C T H is mediated in large part by c A M P [8-10]. Results of some [11-13], but not
There is increasing evidence that serum levels of insulin and dehydroepiandrosterone(sulfate) [DHEA(S)] are functionally related [1-5]. *Present address: Department of Obstetrics and Gynecology, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284-7836. tTo whom correspondence should be addressed: University of Mississippi Medical Center, Department of Pharmacology, 2500 North State Street, Jackson, MS 392164505. Abbreviations: ACTH, adrenocorticotropin,_24; Ang II, [ 5val]angiotensin II; BSA, bovine serum albumin; 8-bromocGMP, 8-bromoguanosine 3': Y-cyclic monophosphate; cAMP, adenosine 3':5' cyclic monophosphate; cptcAMP, 8-(4-chlorophenylthio)-adenosine 3": 5'-cyclic monophosphate; DHEA(S), dehydroepiandrosterone(sulfate); HBSS, Hank's Balanced Salt Solution: Hepes, N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]; IBMX, 3-isobutyl-l-methylxanthine; LDL, low density lipoprotein. 11
12
NANCYA. KLEINet al.
all [14-16], studies suggest that insulin acts to inhibit adenylate cyclase. Moreover, insulin has been reported to increase the activity of cAMP phosphodiesterase [17-19], inhibit cAMPdependent protein kinase[20, 21] and directly oppose the intracellular actions of cAMP [22-24] in a variety of cell types. The present studies were performed to determine if similar actions of insulin in the fasciculata-reticularis cell contribute to the inhibition of ACTHstimulated androstenedione and cortisol secretion. The effects of insulin on the action of ACTH to increase androstenedione and cortisol secretion were compared with its effects on the actions of other factors that initiate steroidogenesis by acting at different sites in the adenylate cyclase/cAMP-dependent protein kinase cascade. The agents selected were forskolin which has been reported to directly activate adenylate cyclase [25] and 8-(4-chlorophenylthio)-cAMP (cpt-cAMP), a phosphodiesterase-resistant cAMP analog that presumably acts at the level of the regulatory subunit of cAMP-dependent protein kinase [for review, see 26]. In addition, the effects of insulin on the steroidogenic response to ACTH were compared with its effects on the steroidogenic response to [Sval]angiotensin II (Ang II), a hormone that is generally acknowledged to act through a calcium-dependent, cAMPindependent mechanism. EXPERIMENTAL
Cell culture Isolated fasciculata-reticularis or glomerulosa cells were prepared from bovine adrenal glands and plated on fibronectin-coated (2#g/cm ~) tissue culture dishes in a modified Ham's F-12/Hepes (25mM, pH 7.4) medium containing 10% horse serum [7, 27]. At confluence, monolayers were incubated for 24-36h in serum- and insulin-free medium supplemented with low density lipoprotein (LDL) (10#g protein/ml) and protease-free bovine serum albumin (BSA) (200/~g/ml). Immediately before use in an experiment, cells were incubated for two 30-45 min periods in Hanks Balanced Salt Solution (HBSS)/Hepes to remove metyrapone present during growth and maintenance periods. Steroid secretion Incubation media were prepared using metyrapone-free modified Ham's F-12/Hepes
medium containing human LDL (10#g/ml), protease-free BSA (200 #g/ml) and bacitracin (200 U/ml). ACTH (Organon Inc., West Orange, NJ, U.S.A.), cpt-cAMP (Sigma, St Louis, MO, U.S.A.), forskolin (Sigma) or Ang II (Bachem Inc., Torrance, CA, U.S.A.) were added at concentrations indicated for specific experiments. Insulin (25 U/mg; Sigma) was added at concentrations of 1.7 × 10-9M (10ng/ml) or 17.5 × 1 0 - 9 M (100 ng/ml); concentrations that produced maximal inhibition of ACTH-stimulated corticosteroid secretion in previous studies [7]. Incubations were allowed to proceed for 24 h. Cortisol and androstenedione contents of the incubation media were determined by radioimmunoassay [28]. Cell protein was measured by the method of Bradford [29]. cAMP production Media of cells incubated for 24h in the absence or presence of ACTH, forskolin, Ang II and/or insulin were removed and replaced with media containing the same experimental regimen and 1 mM IBMX (3-isobutyl-l-methylxanthine), a phosphodiesterase inhibitor. Cells were then incubated for an additional 30 min. The assay was terminated by addition of an equal volume (1 ml) of cold 2 N NaOH. Cells were allowed to stand for 20-30 min and then disrupted by pipeting. An aliquot was removed from each lysate for protein measurement. The remainders were neutralized by addition of concentrated HC1 and diluted as necessary with 0.05 M sodium acetate, cAMP contents of the cell lysates and the media from the preceding 24 h incubation were assayed using a commercially available (Amersham, Arlington Heights, IL, U.S.A.) radioimmunoassay kit. cAMP production in the presence of IBMX was directly dependent upon ACTH concentration and linear over the time employed (data not shown). Degradation of cpt-cAMP Degradation of cpt-cAMP was examined by incubating cells for 24h in the absence or presence of cpt-cAMP (10 3 M) and/or insulin (1.7 or 17.5 × 10 9 M). Incubations were stopped by addition of an equal volume of cold 20% trichloroacetic acid (TCA) containing 8-bromocGMP (2 #g/ml), and the cells were allowed to stand for 30 min at 4°C. Lysates were extracted with chloroform, neutralized by addition of 6 N or 0.6 N KOH, as required. In preparation for HPLC, each lysate was acidified by addition of 1/50th volume 1 N HC1. An aliquot was immedi-
I n s u l i n - A C T H interaction
ately applied to a Waters Ct8 (5/~m) Nova Pak column (0.8 x 10 cm) equilibrated with 20 mM KH2PO 4 (pH 3.7) flowing at a rate of 1.5 ml/min. Cyclic nucleotides were eluted at room temperature using a methanol-H20 (3:2) linear gradient from 0 to 25% over 30 min [30], detected by absorbance at 254 nm, and quantitated by comparison with the absorbances of authentic standards. The column was reequilibrated with 20 mM KH2PO 4 for 15 min between samples. The amounts of cAMP detected in cells incubated in the absence of cpt-cAMP were subtracted from those detected in cells incubated in the presence of cpt-cAMP. Statistics
Cortisol and androstenedione secretion rates were first calculated as nmol x 24 h-l x mg-l cell protein, and secretion rates by control cells or cells treated with insulin alone subtracted from those by treated cells as appropriate. The increases in cortisol and androstenedione production produced by the higher concentration of each agonist (ACTH, forskolin, cpt-cAMP or Ang II) in the absence of insulin were designated as 100%, and the responses by cells treated with the lower concentration of agonist and/or insulin expressed as a percentage of the maximal response. Unless otherwise indicated, 3-4 monolayers were incubated under each condition, and steroid secretion rates averaged to give a single value per experiment. Means of 3-4 experiments were compared by ANOVA, and significant differences between means identified by the methods of least squares and Fisher's least significant difference. Means were considered significantly different at P < 0.05.
13
and cpt-cAMP, respectively. Ang II also caused significant increases in cortisol secretion at concentrations of 10-~0 to 10 -7 M. It is noteworthy, however, that even though cortisol secretion was increased about 7 fold (for example, 3.30 ___ 0.29 nmol x 24 h -~ × mg -1 protein at 10 -s M) by Ang II, the maximal response to Ang II was only a fraction of the response to ACTH. Androstenedione was secreted by control cells at a rate (0.013 ___0.002 nmol x 24h -~ x mg -I protein) that was 25-50 fold lower than the rate of secretion of cortisol. In general, the steroidogenic agonists produced concentrationdependent increases in androstenedione secretion that paralleled those in cortisol secretion [Fig. I(B)]. At the concentrations and duration of treatment used in the present studies, insulin had no effect on basal rates of steroid secretion, and none of the agents, either alone or in combination, affected total cell protein (data not shown). As previously noted [7], insulin attenuated ACTH-stimulated cortisol and androstenedione secretion (Fig. 2). Cortisol secretion [Fig. 2(A)] in response to 10-11M ACTH was reduced 70-80% by insulin (1.7 or 17.5 x 10-gM)
8o
A Cortllol
n, O ,~, 0.5
B
RESULTS
Steroid secretion
The effects of increasing concentrations of ACTH, cpt-cAMP, forskolin or Ang II on cortisol and androstenedione secretion by cultured bovine fasciculata-reticularis cells are presented in Fig. 1. Cortisol secretion [Fig. 1(A)] increased from a basal rate of 0.47+0.02 (mean + SEM) nmol x 24 h-l x mg-~ protein to 67.66 + 7.43 nmol x 24h -1 x mg -~ protein in response to 10-8M ACTH. Comparable increases in cortisol secretion were produced by forskolin (71.02 + 4.96nmol x 24h -~ x mg -~ protein) and cpt-cAMP (54.84_ 3.15 nmol x 24 h-~ x mg-~ protein), but at notably higher concentrations; 10 -5 and 10-3M for forskolin
g
0.3
0.2
~
0.1 0.0
...... i
i
-12
-11
i
-10
i
i
-g
-8
i
i
i
-7
-6
-5
!
-4
i
-3
-2
Iog[AGONIS'r] (M)
Fig. 1. Concentration-dependent effects of steroidogenic agonists on cortisol and androstenedione secretion. Fasciculata-reticularis cells were incubated in the presence of increasing concentrations of A C T H ( Q , 10 -~2 to 10 -7 M), cpt-cAMP ( O , 10 -6 to 2 x 10 -~ M), forskolin (11, 10 -6 to 10 -4 M) or A n g II ( O , 10 -12 to 10 -7 M). Values represent the mean + SEM increase in cortisol (A) or androstenedione (B) secretion rates of 4 individual monolayers from a single cell preparation. The experiment was performed twice with similar results.
NANCY A. KLEIN et al.
14 A
B
100 ~"Corllsol
Androstenedlone
75
1~
Q.
'
1
0 -11
-8
-11
-8
Iog[ACTH] (a) Fig. 2. Effects of insulin on ACTH-stimulated cortisol and androstenedione secretion. Fasciculatareticularis cells were incubated for 24 h in medium with or without A C T H (10 -H or 10 -8 M) in the absence (I-1) or presence of 1.7 × 10 -9 M ([]) or 17.5 x 10 9M ( l ) insulin. The increases in cortisol (A) or androstenedione (B) secretion produced by A C T H under each experimental condition is expressed as a percentage of the cortisol or androstenedione response, respectively, caused by 10 -8 M A C T H alone (100%). Values represent the mean + SEM determined 5-7 separate cell preparations. *P < 0.05 vs 0 insulin at the same cpt-cAMP concentration.
and that in response to 10 -SM ACTH was reduced by about 50%. In general, the effects of insulin on ACTH-stimulated androstenedione secretion [Fig. 2(B)] paralleled its effects on ACTH-stimulated cortisol secretion. Insulin also affected the secretory response to cpt-cAMP (Fig. 3). The effects of insulin on cpt-cAMP-stimulated steroid secretion, however, were not directly comparable to its effects on ACTH-stimulated steroid secretion. The increase in cortisol secretion caused by cpt-cAMP ( 1 0 - 4 o r 10 -3 M ) w a s significantly inhibited only by the higher concentration (17.5 × 10 -9 M ) of insulin [Fig. 3(A)]. Also, the extent of inhibition of cpt-cAMP stimulated cortisol secretion by insulin was about one half that of ACTHstimulated cortisol secretion. The mean rates of cpt-cAMP-stimulated androstenedione secretion were not significantly decreased by insulin, although they were lower in cells incu-
Z
100
~ ~
50
~
2s 0
bated in the presence of both insulin and cptcAMP than in cells incubated in the presence of cpt-cAMP alone [Fig. 3(B)]. The effect of insulin to decrease cpt-cAMP-stimulated cortisol secretion was independent of an effect on the degradation of the cyclic nucleotide. In control cells (4 replicate monolayers) 54.5 + 2.8% of the cpt-cAMP added at the beginning of the experiment was recovered after a 24 h incubation compared to 50.7 ___2.4 and 53.9 + 1.6% in cells incubated in the presence of 1.75 x 10 -9 and 17.5 x 10-9M insulin, respectively. The effects of insulin on forskolin-stimulated cortisol and androstenedione secretion are presented in Fig. 4. At concentrations of 1.7 × 10 -9 and 17.5 x 10-9M, insulin caused 50-70% decreases in cortisol secretion in response to a submaximal concentration ( 1 0 -6 M ) of forskolin [Fig. 4(A)]. Mean rates of androstenedione secretion in response to 10 6 M forskolin
A
B
Cortlsol
Androstenedlone
l
-4
-3
-4
-3
Iog[cpt-cAMPl (M)
Fig. 3. Effects of insulin on cpt-cAMP-stimulated cortisol and androstenedione secretion. Fasciculatareticularis cells were incubated for 24 h in medium with or without cpt-cAMP (I0 4 or 10 -3 M) in the absence (l-q) or presence of 1.7 x 10 -9 M (1~) or 17.5 x 10 -9 M (m) insulin. Values represent the mean _ SEM determined using 3-4 separate cell preparations. Cpt-cAMP-stimulated increases in cortisol (A) and androstenedione (B) secretion are expressed as a percentage of those cells incubated with 10 -3 M cpt-cAMP alone. *P < 0.05 vs 0 insulin at the same cpt-cAMP concentration.
Insulin-ACTH interaction A PJ~H'IQnl
15
B Antlm~mnAtllnna
100 Z
O~ Q
~
og
~ m
7s
~
50
_
~s
o -6
.5
-6
Iog[FORSKOLIN l (M) Fig. 4. Effects of insulin on forskolin-stimulatcd cortiso] and androstenedione secretion. Fasciculatareticularis cells were incubated for 24 h in medium with or without forskolin (10 -6 or 10 -5 M) in the absence (f-l) or presence of 1.7 x 10-gM ([]) or 17.5 x l0 -9 M (11) insulin. Increases in cortisol (A) and androstencdione (B) secretion rates are expressed as a percentage of the cortisol or androstcncdione response of cells incubated with 10 -5 M forskolin alone. Values represent the mean _+ SEM of 3-4 separate cell preparations. *P < 0.05 vs 0 insulin at the same forskolin concentration.
were correspondingly lower in the presence of insulin [Fig. 4(B)], but inhibition was significant only at 17.5 x 10 -9 M insulin. In contrast, cortisol and androstenedione secretion by cells incubated with a maximal concentration (10 -5 M) of forskolin were not consistently affected by insulin. Consequently, even though mean rates of cortisol and androstenedione secretion by cells incubated in the presence of both 10 -5 M forskolin and insulin were some 25% less, they were not significantly different than secretion rates by cells incubated in the presence of forskolin alone. Insulin did not attenuate the increase in cortisol and androstenedione secretion caused by Ang II (Fig. 5). In fact, mean rates of cortisol [Fig. 5(A)] and androstenedione [Fig. 5(B)] secretion in response to Ang II (10 -]l or 10 -~ M) tended to be greater, albeit not significantly, in the presence than in the absence of insulin. The divergence in the actions of insulin on ACTHA
stimulated and Ang II-stimulated steroid secretion observed with fasciculata-reticularis cells was also evident with glomerulosa cells. Aldosterone secretion by glomerulosa cells incubated in the presence of 10 -s M Ang II was not affected by the addition of 17.5 x 10-9M insulin [Fig. 6(B)], while that by cells incubated in the presence of 10 -8 M A C T H was attenuated by 50-60% [Fig. 6(C)]. cAMP
The effects of insulin on the amounts of cAMP accumulated in the media of fasciculatareticularis cells during a 24 h incubation in the presence of maximal concentrations of agonist roughly paralleled its effects on steroid secretion (Table 1). cAMP content increased approx. 8 fold in response to 10-8M A C T H and 5 fold in response to 10-SM forskolin. In cells incubated in the presence of l0 8M Ang II, the amount of cAMP accumulated in the medium B
2O0
Z
o~ ~. 150 0 OZ
100
~
so
o
-11
-8
-11
-8
Iog[ANG II] (M) Fig. 5. Effects of insulin on Ang II-stimulated cortisol and androstenedione secretion. Fasciculatareticularis cells were incubated for 24h in medium with or without Ang II (I0 -t~ or 10-SM) in the absence (VI) or presence of 1.7 x 10 -9 M ([]) or 17.5 x 10 -9 M ( I ) insulin. The increase in cortisol (A) and androstenedione (B) secretion evoked by each treatment regimen is expressed as a percentage of the increase in cortisol or androstenedione secretion produced by 10 -s M Ang II alone. Values represent the mean + SEM of 3-4 separate cell preparations. *P < 0.05 vs 0 insulin at the same forskolin concentration.
NANCY A. KLEIN et al.
16
Table 1. Effects of insulin on agonist-stimulated cAMP accumulation Insulin (10 -9 M) Agonist None A C T H (10 -8 M) (10 11M) Forskolin (10 5M) (10-rM) Ang II (10-8 M)
0
1.75
17.5
79.6 ± 5. l 529.7 ± 24.7* 146.8 ± 6.8* 353.5±62.1 145.1±1.6" 108.8 ± 14.1
392.8 ± 13.7"* 112.7 + 5.6** 362.4±61.2 130.8_+8.8 104.9_+ 13.0
319.6 ± 18.0"* 122.0 ± 4.6** 346.2 ± 11.8 100.4_+6.1"* 138.5± 14.2
Fasciculata-reticularis cells were incubated for 24 h in the absence (none) or presence of ACTH, forskolin or Ang II alone or in combination with insulin, cAMP accumulation is expressed as pmol x 24 h × mg t cell protein. Values represent the mean ± SEM of 3 replicate monolayers from a single cell preparation. The experiment was performed twice with similar results. *P < 0.01 vs "none", **P ~
0960-0760/92 $5.00 + 0.00 Copyright © 1992 Pergamon Press pie
Printed in Great Britain. All rights reserved
MECHANISMS OF INSULIN INHIBITION OF ACTH-STIMULATED STEROID SECRETION BY CULTURED BOVINE ADRENOCORTICAL CELLS NANCY A. KLEIN, 1. RICHARD N. ANDERSEN,1 PETER R. CASSON,1 JOHN E. BUSTERl
and ROBERTE. KRAMER2t Departments of ~Obstetrics and Gynecology and 2Pharmacology, University of Tennessee, Memphis, TN 38163, U.S.A. (Received 14 March 1991)
Summary--Results of previous studies indicated that insulin at levels comparable to those in humans during hyperinsulinemia decreased ACTH-stimulated cortisol and androstenedione secretion by bovine adrenal fasciculata-reticularis cells in primary culture. In the present studies this inhibitory action was examined further by comparing the effects of insulin on ACTH-stimulated corticosteroid secretion with its effects on 8-(4-chlorophenylthio)-cAMP (cpt-cAMP), forskolin- and [Sval]angiotensin II (Ang II)-stimulated corticosteroid secretion. Effects on corticosteroid secretion were correlated with effects on cAMP accumulation and rates of cAMP production. Monolayers were incubated for 24 h in the absence or presence of each agonist alone or in combination with insulin. Insulin (1.7 x 10-9 or 17.5 x 10 -9 M) caused about a 50% decrease in cortisol and androstenedione secretion in response to ACTH (10-'1 or 10-s M). Insulin also decreased ACTH-stimulated aldosterone secretion by cultured glomerulosa cells. Cpt-cAMP (10 -4 or 10 -3 M)-stimulated increases in cortisol and androstenedione secretion were inhibited by insulin, but to a lesser extent than those in response to ACTH. The inhibition of cpt-cAMP-stimulated steroid secretion was not related to increased degradation of the cyclic nucleotide. Increases in cortisol and androstenedione secretion caused by a submaximal concentration (10-6 M) of forskolin were decreased 50-70% by insulin. In contrast, insulin failed to significantly affect cortisol or androstenedione secretion caused by a maximal concentration (10 -5 M) of forskolin. The secretory responses to Ang II (10 -s M) were also unaffected by insulin. The effect of insulin to inhibit ACTH-stimulated steroid secretion was accompanied by a reduction in cAMP accumulation as well as an apparent inhibition of adenylate cyclase activation. These data indicate that the effect of insulin to attenuate ACTH-stimulated corticosteroid secretion results from both an inhibition of ACTH-stimulated adenylate cyclase activity and an antagonism of the intracellular actions of cAMP.
INTRODUCTION
Studies by Farah et al. [6] have extended these observations and demonstrated bimodal effects of insulin on D H E A production in vivo, Farah et al's findings are in agreement with the results of recent studies [7] from this laboratory in which ACTH-stimulated androstenedione secretion by primary cultures of bovine fasciculata-reticularis cells was stimulated by insulin at physiological levels (~< 1.7 pM) and inhibited by insulin at hyperinsulinemic levels (1> 1.7 nM). High concentrations of insulin also inhibited ACTH-stimulated cortisol secretion. The present study was undertaken to identify the mechanism(s) underlying the inhibitory action of insulin on ACTH-stimulated androstenedione and cortisol secretion. It is generally accepted that the steroidogenic effect of A C T H is mediated in large part by c A M P [8-10]. Results of some [11-13], but not
There is increasing evidence that serum levels of insulin and dehydroepiandrosterone(sulfate) [DHEA(S)] are functionally related [1-5]. *Present address: Department of Obstetrics and Gynecology, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284-7836. tTo whom correspondence should be addressed: University of Mississippi Medical Center, Department of Pharmacology, 2500 North State Street, Jackson, MS 392164505. Abbreviations: ACTH, adrenocorticotropin,_24; Ang II, [ 5val]angiotensin II; BSA, bovine serum albumin; 8-bromocGMP, 8-bromoguanosine 3': Y-cyclic monophosphate; cAMP, adenosine 3':5' cyclic monophosphate; cptcAMP, 8-(4-chlorophenylthio)-adenosine 3": 5'-cyclic monophosphate; DHEA(S), dehydroepiandrosterone(sulfate); HBSS, Hank's Balanced Salt Solution: Hepes, N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]; IBMX, 3-isobutyl-l-methylxanthine; LDL, low density lipoprotein. 11
12
NANCYA. KLEINet al.
all [14-16], studies suggest that insulin acts to inhibit adenylate cyclase. Moreover, insulin has been reported to increase the activity of cAMP phosphodiesterase [17-19], inhibit cAMPdependent protein kinase[20, 21] and directly oppose the intracellular actions of cAMP [22-24] in a variety of cell types. The present studies were performed to determine if similar actions of insulin in the fasciculata-reticularis cell contribute to the inhibition of ACTHstimulated androstenedione and cortisol secretion. The effects of insulin on the action of ACTH to increase androstenedione and cortisol secretion were compared with its effects on the actions of other factors that initiate steroidogenesis by acting at different sites in the adenylate cyclase/cAMP-dependent protein kinase cascade. The agents selected were forskolin which has been reported to directly activate adenylate cyclase [25] and 8-(4-chlorophenylthio)-cAMP (cpt-cAMP), a phosphodiesterase-resistant cAMP analog that presumably acts at the level of the regulatory subunit of cAMP-dependent protein kinase [for review, see 26]. In addition, the effects of insulin on the steroidogenic response to ACTH were compared with its effects on the steroidogenic response to [Sval]angiotensin II (Ang II), a hormone that is generally acknowledged to act through a calcium-dependent, cAMPindependent mechanism. EXPERIMENTAL
Cell culture Isolated fasciculata-reticularis or glomerulosa cells were prepared from bovine adrenal glands and plated on fibronectin-coated (2#g/cm ~) tissue culture dishes in a modified Ham's F-12/Hepes (25mM, pH 7.4) medium containing 10% horse serum [7, 27]. At confluence, monolayers were incubated for 24-36h in serum- and insulin-free medium supplemented with low density lipoprotein (LDL) (10#g protein/ml) and protease-free bovine serum albumin (BSA) (200/~g/ml). Immediately before use in an experiment, cells were incubated for two 30-45 min periods in Hanks Balanced Salt Solution (HBSS)/Hepes to remove metyrapone present during growth and maintenance periods. Steroid secretion Incubation media were prepared using metyrapone-free modified Ham's F-12/Hepes
medium containing human LDL (10#g/ml), protease-free BSA (200 #g/ml) and bacitracin (200 U/ml). ACTH (Organon Inc., West Orange, NJ, U.S.A.), cpt-cAMP (Sigma, St Louis, MO, U.S.A.), forskolin (Sigma) or Ang II (Bachem Inc., Torrance, CA, U.S.A.) were added at concentrations indicated for specific experiments. Insulin (25 U/mg; Sigma) was added at concentrations of 1.7 × 10-9M (10ng/ml) or 17.5 × 1 0 - 9 M (100 ng/ml); concentrations that produced maximal inhibition of ACTH-stimulated corticosteroid secretion in previous studies [7]. Incubations were allowed to proceed for 24 h. Cortisol and androstenedione contents of the incubation media were determined by radioimmunoassay [28]. Cell protein was measured by the method of Bradford [29]. cAMP production Media of cells incubated for 24h in the absence or presence of ACTH, forskolin, Ang II and/or insulin were removed and replaced with media containing the same experimental regimen and 1 mM IBMX (3-isobutyl-l-methylxanthine), a phosphodiesterase inhibitor. Cells were then incubated for an additional 30 min. The assay was terminated by addition of an equal volume (1 ml) of cold 2 N NaOH. Cells were allowed to stand for 20-30 min and then disrupted by pipeting. An aliquot was removed from each lysate for protein measurement. The remainders were neutralized by addition of concentrated HC1 and diluted as necessary with 0.05 M sodium acetate, cAMP contents of the cell lysates and the media from the preceding 24 h incubation were assayed using a commercially available (Amersham, Arlington Heights, IL, U.S.A.) radioimmunoassay kit. cAMP production in the presence of IBMX was directly dependent upon ACTH concentration and linear over the time employed (data not shown). Degradation of cpt-cAMP Degradation of cpt-cAMP was examined by incubating cells for 24h in the absence or presence of cpt-cAMP (10 3 M) and/or insulin (1.7 or 17.5 × 10 9 M). Incubations were stopped by addition of an equal volume of cold 20% trichloroacetic acid (TCA) containing 8-bromocGMP (2 #g/ml), and the cells were allowed to stand for 30 min at 4°C. Lysates were extracted with chloroform, neutralized by addition of 6 N or 0.6 N KOH, as required. In preparation for HPLC, each lysate was acidified by addition of 1/50th volume 1 N HC1. An aliquot was immedi-
I n s u l i n - A C T H interaction
ately applied to a Waters Ct8 (5/~m) Nova Pak column (0.8 x 10 cm) equilibrated with 20 mM KH2PO 4 (pH 3.7) flowing at a rate of 1.5 ml/min. Cyclic nucleotides were eluted at room temperature using a methanol-H20 (3:2) linear gradient from 0 to 25% over 30 min [30], detected by absorbance at 254 nm, and quantitated by comparison with the absorbances of authentic standards. The column was reequilibrated with 20 mM KH2PO 4 for 15 min between samples. The amounts of cAMP detected in cells incubated in the absence of cpt-cAMP were subtracted from those detected in cells incubated in the presence of cpt-cAMP. Statistics
Cortisol and androstenedione secretion rates were first calculated as nmol x 24 h-l x mg-l cell protein, and secretion rates by control cells or cells treated with insulin alone subtracted from those by treated cells as appropriate. The increases in cortisol and androstenedione production produced by the higher concentration of each agonist (ACTH, forskolin, cpt-cAMP or Ang II) in the absence of insulin were designated as 100%, and the responses by cells treated with the lower concentration of agonist and/or insulin expressed as a percentage of the maximal response. Unless otherwise indicated, 3-4 monolayers were incubated under each condition, and steroid secretion rates averaged to give a single value per experiment. Means of 3-4 experiments were compared by ANOVA, and significant differences between means identified by the methods of least squares and Fisher's least significant difference. Means were considered significantly different at P < 0.05.
13
and cpt-cAMP, respectively. Ang II also caused significant increases in cortisol secretion at concentrations of 10-~0 to 10 -7 M. It is noteworthy, however, that even though cortisol secretion was increased about 7 fold (for example, 3.30 ___ 0.29 nmol x 24 h -~ × mg -1 protein at 10 -s M) by Ang II, the maximal response to Ang II was only a fraction of the response to ACTH. Androstenedione was secreted by control cells at a rate (0.013 ___0.002 nmol x 24h -~ x mg -I protein) that was 25-50 fold lower than the rate of secretion of cortisol. In general, the steroidogenic agonists produced concentrationdependent increases in androstenedione secretion that paralleled those in cortisol secretion [Fig. I(B)]. At the concentrations and duration of treatment used in the present studies, insulin had no effect on basal rates of steroid secretion, and none of the agents, either alone or in combination, affected total cell protein (data not shown). As previously noted [7], insulin attenuated ACTH-stimulated cortisol and androstenedione secretion (Fig. 2). Cortisol secretion [Fig. 2(A)] in response to 10-11M ACTH was reduced 70-80% by insulin (1.7 or 17.5 x 10-gM)
8o
A Cortllol
n, O ,~, 0.5
B
RESULTS
Steroid secretion
The effects of increasing concentrations of ACTH, cpt-cAMP, forskolin or Ang II on cortisol and androstenedione secretion by cultured bovine fasciculata-reticularis cells are presented in Fig. 1. Cortisol secretion [Fig. 1(A)] increased from a basal rate of 0.47+0.02 (mean + SEM) nmol x 24 h-l x mg-~ protein to 67.66 + 7.43 nmol x 24h -1 x mg -~ protein in response to 10-8M ACTH. Comparable increases in cortisol secretion were produced by forskolin (71.02 + 4.96nmol x 24h -~ x mg -~ protein) and cpt-cAMP (54.84_ 3.15 nmol x 24 h-~ x mg-~ protein), but at notably higher concentrations; 10 -5 and 10-3M for forskolin
g
0.3
0.2
~
0.1 0.0
...... i
i
-12
-11
i
-10
i
i
-g
-8
i
i
i
-7
-6
-5
!
-4
i
-3
-2
Iog[AGONIS'r] (M)
Fig. 1. Concentration-dependent effects of steroidogenic agonists on cortisol and androstenedione secretion. Fasciculata-reticularis cells were incubated in the presence of increasing concentrations of A C T H ( Q , 10 -~2 to 10 -7 M), cpt-cAMP ( O , 10 -6 to 2 x 10 -~ M), forskolin (11, 10 -6 to 10 -4 M) or A n g II ( O , 10 -12 to 10 -7 M). Values represent the mean + SEM increase in cortisol (A) or androstenedione (B) secretion rates of 4 individual monolayers from a single cell preparation. The experiment was performed twice with similar results.
NANCY A. KLEIN et al.
14 A
B
100 ~"Corllsol
Androstenedlone
75
1~
Q.
'
1
0 -11
-8
-11
-8
Iog[ACTH] (a) Fig. 2. Effects of insulin on ACTH-stimulated cortisol and androstenedione secretion. Fasciculatareticularis cells were incubated for 24 h in medium with or without A C T H (10 -H or 10 -8 M) in the absence (I-1) or presence of 1.7 × 10 -9 M ([]) or 17.5 x 10 9M ( l ) insulin. The increases in cortisol (A) or androstenedione (B) secretion produced by A C T H under each experimental condition is expressed as a percentage of the cortisol or androstenedione response, respectively, caused by 10 -8 M A C T H alone (100%). Values represent the mean + SEM determined 5-7 separate cell preparations. *P < 0.05 vs 0 insulin at the same cpt-cAMP concentration.
and that in response to 10 -SM ACTH was reduced by about 50%. In general, the effects of insulin on ACTH-stimulated androstenedione secretion [Fig. 2(B)] paralleled its effects on ACTH-stimulated cortisol secretion. Insulin also affected the secretory response to cpt-cAMP (Fig. 3). The effects of insulin on cpt-cAMP-stimulated steroid secretion, however, were not directly comparable to its effects on ACTH-stimulated steroid secretion. The increase in cortisol secretion caused by cpt-cAMP ( 1 0 - 4 o r 10 -3 M ) w a s significantly inhibited only by the higher concentration (17.5 × 10 -9 M ) of insulin [Fig. 3(A)]. Also, the extent of inhibition of cpt-cAMP stimulated cortisol secretion by insulin was about one half that of ACTHstimulated cortisol secretion. The mean rates of cpt-cAMP-stimulated androstenedione secretion were not significantly decreased by insulin, although they were lower in cells incu-
Z
100
~ ~
50
~
2s 0
bated in the presence of both insulin and cptcAMP than in cells incubated in the presence of cpt-cAMP alone [Fig. 3(B)]. The effect of insulin to decrease cpt-cAMP-stimulated cortisol secretion was independent of an effect on the degradation of the cyclic nucleotide. In control cells (4 replicate monolayers) 54.5 + 2.8% of the cpt-cAMP added at the beginning of the experiment was recovered after a 24 h incubation compared to 50.7 ___2.4 and 53.9 + 1.6% in cells incubated in the presence of 1.75 x 10 -9 and 17.5 x 10-9M insulin, respectively. The effects of insulin on forskolin-stimulated cortisol and androstenedione secretion are presented in Fig. 4. At concentrations of 1.7 × 10 -9 and 17.5 x 10-9M, insulin caused 50-70% decreases in cortisol secretion in response to a submaximal concentration ( 1 0 -6 M ) of forskolin [Fig. 4(A)]. Mean rates of androstenedione secretion in response to 10 6 M forskolin
A
B
Cortlsol
Androstenedlone
l
-4
-3
-4
-3
Iog[cpt-cAMPl (M)
Fig. 3. Effects of insulin on cpt-cAMP-stimulated cortisol and androstenedione secretion. Fasciculatareticularis cells were incubated for 24 h in medium with or without cpt-cAMP (I0 4 or 10 -3 M) in the absence (l-q) or presence of 1.7 x 10 -9 M (1~) or 17.5 x 10 -9 M (m) insulin. Values represent the mean _ SEM determined using 3-4 separate cell preparations. Cpt-cAMP-stimulated increases in cortisol (A) and androstenedione (B) secretion are expressed as a percentage of those cells incubated with 10 -3 M cpt-cAMP alone. *P < 0.05 vs 0 insulin at the same cpt-cAMP concentration.
Insulin-ACTH interaction A PJ~H'IQnl
15
B Antlm~mnAtllnna
100 Z
O~ Q
~
og
~ m
7s
~
50
_
~s
o -6
.5
-6
Iog[FORSKOLIN l (M) Fig. 4. Effects of insulin on forskolin-stimulatcd cortiso] and androstenedione secretion. Fasciculatareticularis cells were incubated for 24 h in medium with or without forskolin (10 -6 or 10 -5 M) in the absence (f-l) or presence of 1.7 x 10-gM ([]) or 17.5 x l0 -9 M (11) insulin. Increases in cortisol (A) and androstencdione (B) secretion rates are expressed as a percentage of the cortisol or androstcncdione response of cells incubated with 10 -5 M forskolin alone. Values represent the mean _+ SEM of 3-4 separate cell preparations. *P < 0.05 vs 0 insulin at the same forskolin concentration.
were correspondingly lower in the presence of insulin [Fig. 4(B)], but inhibition was significant only at 17.5 x 10 -9 M insulin. In contrast, cortisol and androstenedione secretion by cells incubated with a maximal concentration (10 -5 M) of forskolin were not consistently affected by insulin. Consequently, even though mean rates of cortisol and androstenedione secretion by cells incubated in the presence of both 10 -5 M forskolin and insulin were some 25% less, they were not significantly different than secretion rates by cells incubated in the presence of forskolin alone. Insulin did not attenuate the increase in cortisol and androstenedione secretion caused by Ang II (Fig. 5). In fact, mean rates of cortisol [Fig. 5(A)] and androstenedione [Fig. 5(B)] secretion in response to Ang II (10 -]l or 10 -~ M) tended to be greater, albeit not significantly, in the presence than in the absence of insulin. The divergence in the actions of insulin on ACTHA
stimulated and Ang II-stimulated steroid secretion observed with fasciculata-reticularis cells was also evident with glomerulosa cells. Aldosterone secretion by glomerulosa cells incubated in the presence of 10 -s M Ang II was not affected by the addition of 17.5 x 10-9M insulin [Fig. 6(B)], while that by cells incubated in the presence of 10 -8 M A C T H was attenuated by 50-60% [Fig. 6(C)]. cAMP
The effects of insulin on the amounts of cAMP accumulated in the media of fasciculatareticularis cells during a 24 h incubation in the presence of maximal concentrations of agonist roughly paralleled its effects on steroid secretion (Table 1). cAMP content increased approx. 8 fold in response to 10-8M A C T H and 5 fold in response to 10-SM forskolin. In cells incubated in the presence of l0 8M Ang II, the amount of cAMP accumulated in the medium B
2O0
Z
o~ ~. 150 0 OZ
100
~
so
o
-11
-8
-11
-8
Iog[ANG II] (M) Fig. 5. Effects of insulin on Ang II-stimulated cortisol and androstenedione secretion. Fasciculatareticularis cells were incubated for 24h in medium with or without Ang II (I0 -t~ or 10-SM) in the absence (VI) or presence of 1.7 x 10 -9 M ([]) or 17.5 x 10 -9 M ( I ) insulin. The increase in cortisol (A) and androstenedione (B) secretion evoked by each treatment regimen is expressed as a percentage of the increase in cortisol or androstenedione secretion produced by 10 -s M Ang II alone. Values represent the mean + SEM of 3-4 separate cell preparations. *P < 0.05 vs 0 insulin at the same forskolin concentration.
NANCY A. KLEIN et al.
16
Table 1. Effects of insulin on agonist-stimulated cAMP accumulation Insulin (10 -9 M) Agonist None A C T H (10 -8 M) (10 11M) Forskolin (10 5M) (10-rM) Ang II (10-8 M)
0
1.75
17.5
79.6 ± 5. l 529.7 ± 24.7* 146.8 ± 6.8* 353.5±62.1 145.1±1.6" 108.8 ± 14.1
392.8 ± 13.7"* 112.7 + 5.6** 362.4±61.2 130.8_+8.8 104.9_+ 13.0
319.6 ± 18.0"* 122.0 ± 4.6** 346.2 ± 11.8 100.4_+6.1"* 138.5± 14.2
Fasciculata-reticularis cells were incubated for 24 h in the absence (none) or presence of ACTH, forskolin or Ang II alone or in combination with insulin, cAMP accumulation is expressed as pmol x 24 h × mg t cell protein. Values represent the mean ± SEM of 3 replicate monolayers from a single cell preparation. The experiment was performed twice with similar results. *P < 0.01 vs "none", **P ~
