Comparative Effect of Pituitary Adenylate Cyclase- Activating Polypeptide on Aldosterone Secretion in Normal Bovine and Human Tumorous Adrenal Cells*
V. BODART, K. BABINSKI, H. ONG, AND A. DE LÉANÝ
Faculty of Pharmacy, Department of Pharmacology, Faculty of Medicine, Université de Montréal, Montréal, Québec H3C 3J7, Canada
ABSTRACT
The purpose of this study was to investigate the mechanisms of action of pituitary adenylate cyclase-activating polypeptide (PACAP) in stimulating aldosterone production in two different models: bovine adrenal zona glomerulosa (ZG) cells in primary culture and the hu- man adrenocortical carcinoma cell line H295R. PACAP binds to two major groups of receptors: type I, which prefers PACAP38 and PACAP27 over vasoactive intestinal peptide (VIP); and type II, which has approximately equal affinity for PACAP38, PACAP27, and VIP. The type I subclass comprises multiple splice variants that can be distinguished by their specificity to PACAP38 and PACAP27 in their activation of adenylate cyclase and phospholipase C. Type II PACAP/ VIP receptors couple only to AC. In bovine ZG cells, PACAP38 and PACAP27 stimulated aldosterone production in a dose-dependent manner, whereas VIP was ineffective. In H295R cells, PACAP38, PACAP27, and VIP dose-dependently stimulated aldosterone produc-
tion with roughly the same ED50. In bovine ZG cells, PACAP38 and PACAP27 stimulated cAMP production with similar efficacy, whereas VIP had no effect. In H295R cells, all three peptides stimulated cAMP accumulation. PACAP38 and PACAP27 also activated PLC in bovine ZG cells as they induced an increase in Ins(1,4,5)P3 production. In H295R cells, neither of these peptides was able to stimulate IP turn- over. These results indicate that PACAP stimulation of aldosterone production is mediated by the PVR1s or the PVR1hop splice variants of the type I PACAP-specific receptor subtype in bovine ZG cells, whereas only type II PACAP/VIP receptors seemed to occur in the human H295R cell line. In addition, PACAP-stimulated aldosterone production was inhibited by atrial natriuretic peptide in bovine and human adrenocortical cells, however not by the same mechanism. This further supports species-specific and/or cell type-specific signal- ing pathways for PACAP in the regulation of aldosterone production. (Endocrinology 138: 566-573, 1997)
P ITUITARY adenylate cyclase-activating polypeptide (PACAP) was first isolated from ovine hypothalamus on the basis of its potent activity in stimulating cAMP pro- duction in rat anterior pituitary cells. Considering its high level of sequence homology with vasoactive intestinal pep- tide (VIP), PACAP has been classified as a member of the secretin/glucagon/VIP polypeptide family. The PACAP precursor protein is processed into two C-terminally ami- dated peptides: PACAP38, with 38 amino acids; and PACAP27, corresponding to the 27 N-terminal amino acids of PACAP38. PACAP27 and PACAP38 are both present in the central nervous system, as well as in the periphery, al- though PACAP27 represents only a minor portion of total PACAP immunoreactivity in many tissues (1, 2).
To date, three receptors for PACAP have been cloned and are distinguished on a pharmacological basis by their relative affinity for PACAP and VIP and by the transduction path- ways these peptides are able to activate. The type I receptor (PVR1) binds PACAP 100- to 1000-fold more potently than VIP and is coupled through G proteins to the activation of both adenylate cyclase (AC) and phospholipase C (PLC).
Alternative splicing of this receptor may change the signal transduction characteristics by altering the structure of the third cytoplasmic loop putatively involved in G protein cou- pling. Type II receptors (PVR2 and PVR3) bind PACAP and VIP with approximately equal affinities. These receptors are coupled, probably through the G protein Gs, to the activation of AC (1, 2).
PACAP-like immunoreactivity and PACAP-containing nerve fibers have been demonstrated in the adrenal gland of several species (frog, mouse, hamster, rat, cow, and pig) (3), as well as in human pheochromocytomas (4). The occurrence of specific binding sites for PACAP has been shown in rat adrenal medulla (5) and on both adrenocortical and chro- maffin cells of frog adrenal gland (6). In this later species, PACAP has been found to directly stimulate aldosterone and corticosterone secretion by dispersed, perifused adrenal cells. In rats and humans, PACAP stimulated aldosterone and corticosterone/cortisol secretion by adrenal slices, but not by dispersed adrenocortical cells, suggesting a rather indirect role of PACAP mediated by the chromaffin cells of the adrenal medulla (7, 8). In vivo studies have shown that PACAP increases the mean plasma cortisol concentration in conscious, functionally hypophysectomized calves (9).
The present study was designed to investigate the poten- tial direct stimulatory effect of PACAP on aldosterone syn- thesis and secretion in mammalian species. Because respon- siveness of a cell to PACAP will depend upon the type(s) of PVR expressed and the types of intracellular signaling path- way involved, it was tempting to compare the effects of
Address all correspondence and requests for reprints to: Dr. A. De Léan, Department of Pharmacology, Faculty of Medicine, Université de Montréal, Case Postale 6128, Succursale Centre-Ville, Montréal H3C 3J7, Canada. E-mail: delean@ere.umontreal.ca.
* This work was supported by a program grant from the Medical Research Council of Canada.
+ Recipient of a PMAC-MRC research chair in Pharmacology spon- sored by Merck-Frosst Canada.
PACAP in two different cell system models. The two models selected for our study are the bovine adrenal zona glomeru- losa (ZG) cells in primary culture and the human adreno- cortical carcinoma cell line H295R, which has been previ- ously shown to retain functional control of aldosterone secretion by angiotensin II (AII) and atrial natriuretic peptide (ANP) (10,11). The effects of PACAP and VIP on aldosterone, cAMP, and inositol(1,4,5)trisphosphate (Ins(1,4,5) P3) pro- ductions were studied in bovine and human adrenal cells. The effect of PACAP was compared with that of AII in both cell systems. Finally, the ability of ANP to inhibit PACAP- stimulated aldosterone synthesis was determined because this peptide is known as the physiological inhibitor of aldo- sterone synthesis and capable of counteracting the effects of most, if not all, secretagogues of this steroid hormone (12).
Materials and Methods
Materials
Ham’s F12 medium, Dulbecco’s modified Eagle’s-Ham’s F-12 me- dium (DMEM/F12), horse serum, FBS, and antibiotics were purchased from GIBCO Labs Inc. (Burlington, Ontario). Nu-Serum serum replace- ment and ITS+ Premix Universal Culture supplement were from Col- laborative Biomedical Products Inc. (Bedford, MA). Collagenase type IA, DNAse type I, protein kinase (PKA), cAMP-dependent protein kinase from porcine heart and cAMP were from Sigma Chemical Co. (St. Louis, MO). AII and rat ANP were from Peninsula (Belmont, CA). PACAP38 (human, ovine, rat), PACAP27 (human, ovine, rat), and VIP (human, porcine, rat) were purchased from Phoenix Pharmaceuticals, Inc. (Bel- mont, CA). [8-3H]CAMP, D-myo-[3H]Inositol 1, 4, 5-trisphosphate and D-myo-Inositol 1, 4, 5-trisphosphate [Ins(1, 4, 5)P3] were from Amersham (Oakville, Ontario). Aldosterone-3-(O-carboxymethyl)oximino-(2- [125I]iodohistamine) was purchased from Diagnostic Products Corpo- ration (Markham, Ontario). Anti-aldosterone-3-BSA-antibody was from ICN Biomedicals Inc. (Costa Mesa, CA).
Cell culture
Primary culture of bovine adrenal ZG cells was performed as de- scribed (13). Briefly, bovine adrenal glands were obtained from a local slaughterhouse. The glands were cleaned of fat and a 0.5-mm layer containing the capsule and the ZG was dissected with a scalpel. The cells were dispersed in Ham’s F12 medium with 0.2% collagenase type IA, 0.025% DNAse type I, and 0.25% BSA. Washed cells were resuspended in Ham’s F12 medium supplemented with 10% horse serum, 2% FBS, 1% streptomycin, 1% penicillin, and 2.5 mg/ml fungizone. The cell sus- pension (106 cells/ml) was distributed in 1-ml fractions in 24-well cluster plates for aldosterone and cAMP determinations and in 5-ml fractions in 6-well plates for Ins(1, 4, 5)P3 determinations. Cell viability, as mon- itored by trypan blue exclusion, was generally greater than 95%. Con- tamination of the ZG cell preparations by zona fasciculata cells was less than 10% as attested by morphological observation of the cells in the light microscope and measurement of cortisol production by these cell prep- arations. The cells were cultured in serum-containing medium for 4 days and starved in serum-free medium for 48 h. The viability of the cells was not affected during the 6 days of culture or by the various cell treatments.
H295R cells were selected from the NCI-H295 cell line obtained from the American Type Culture Collection (ATCC, Rockville, MD) as pre- viously reported (10). The cells were maintained in a 1:1 mixture of DMEM/F12 containing pyridoxine HCI, L-glutamine, and 15 mM HEPES and supplemented with insulin, transferrin, selenium (1% ITS+), 2.5% Nu-Serum, and antibiotics. Cells were grown in 75-cm2 flasks at 37 C under an atmosphere of 5% CO2-95% air until they reached conflu- ence. Where aldosterone and cAMP production were studied, confluent cell monolayers were subcultured in 24-well cluster plates, and after 48 h, medium was replaced with fresh serum-free medium (DMEM/F12 containing 0.01% BSA). Cells were cultured further for 24 h, then rinsed and treated in the same medium. For Ins(1,4,5)P3 measurements, H295R cells were subcultured in 6-well plates.
Aldosterone determination
The cells were washed with their respective medium without serum, and quadruplicate cell culture wells were stimulated for 3 h at 37 C with various concentrations of PACAP and VIP added to fresh serum-free medium containing 0.01% BSA. At the end of the incubation, the me- dium was removed quickly and frozen at -20 C until assayed for aldosterone determination. Aldosterone was directly measured in cell culture medium by a specific RIA as already described (14). The least detectable concentration measured by the RIA was 5 fmol/ml.
cAMP determination
Bovine ZG cells and H295R cells were washed with their respective serum-free medium and preincubated in 0.5 ml of the same medium containing 0.01% BSA and 0.5 mm of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX) for 20 min at 37 C. After the pre- incubation period, the cells were incubated for 20 min at 37 C in the same medium containing various concentrations of PACAP and VIP. At the end of the incubation, intracellular cAMP content was extracted with 0.5 ml of 100% ethanol/10 mM HC1. Cell extracts were evaporated to dry- ness in a Speedvac concentrator (Savant) and resuspended in 50 pl of binding buffer.
cAMP was measured by a specific competitive binding assay using cAMP-dependent PKA from porcine heart as binding protein (15).The binding buffer consisted of 50 mM Tris-HC1 pH 7.4 and 4 mM EDTA. Assays were conducted in 1.5-ml Eppendorff tubes and each contained 50 pl of either cAMP standard (0-100 pM) or cell extract, 50 ul of [3 H]-CAMP (approximately 50,000 cpm) and 100 pl of PKA (4 µg, diluted in binding buffer plus 0.1% BSA). The tubes were incubated on ice for 2 h. The reaction was terminated by adding 100 ul of 3.5% (wt/vol) charcoal in binding buffer plus 2% BSA. The tubes were briefly agitated, centrifugated (12,000 g, 5 min, 4 C) and 200 ul of the supernatant were taken for liquid scintillation counting. The lower limit of detection of this assay was 0.140 pmol cAMP per tube.
Inositol (1, 4, 5) trisphosphate determination
H295R cells and bovine ZG cells were washed twice and preincubated 30 min at 37 C in incubation buffer (145 mM NaCl, 5.6 mm KC1, 5.6 mM glucose, 0.01% BSA, 10 mM HEPES pH 7.4). After the preincubation period, the cells were stimulated with PACAP and VIP in 1 ml of warmed incubation buffer for 10 sec. The incubations were terminated by addition of 250 ul ice-cold perchloric acid (10%, vol/vol). The cells were scraped with a rubber policeman and each well was washed with 250 µ1 10% perchloric acid. Samples were centrifugated (12,000 × g, 5 min, 4 C), and the supernatants were neutralized with 1.5 M KOH containing 60 mM HEPES in the presence of Universal Indicator.
Ins(1,4,5)P3 was quantified by a specific competitive binding assay using a crude microsomal preparation of bovine adrenal cortex (16). The binding assay was conducted in 1.5-ml Eppendorff tubes. Each incu- bation contained 25 ul of incubation buffer (100 mM Tris-HC1 pH 9.0, 4 mM EDTA, 4 mM EGTA, 0.4% (wt/vol) BSA), 25 pl [3 H]-Ins(1, 4, 5)P3 (approximately 3000 cpm), 25 pl of standard Ins(1, 4, 5)P3 (0-100 pM) or cell extract and 25 ul of binding protein (0.5 mg). The tubes were agitated and incubated for 45 min on ice. Incubations were terminated by cen- trifugation (12,000 × g, 5 min, 4 C). The supernatant was removed by aspiration, and the pellet was dissolved in scintillation liquid and counted. The lower limit of detection of this assay was 0.200 pmol Ins(1, 4, 5)P3 per tube.
Data analysis
Dose-response curves for aldosterone and cAMP production were analyzed with the ALLFIT for Windows1 program based on a four- parameter logistic equation to obtain estimates of the ED50 (17). Statis- tical differences were tested by ANOVA and Bonferroni multiple com- parison t test after checking for homogeneity of variance. Statistical significance levels were set at P < 0.05 (significant) and P < 0.01 (highly
1 Requests for the program ALLFIT for Windows can be addressed by E-mail to: delean@ere.umontreal.ca.
significant). All data presented are representative of at least two separate experiments.
Results
Effects of PACAP and VIP on aldosterone production in human and bovine adrenal cells
Treatment of adrenal cortical cells with PACAP dem- onstrated that this neuropeptide could directly stimulate aldosterone production in both human and bovine species (Fig. 1). PACAP acts on target cells by binding to cognate receptors. These receptors are divided in two subtypes, each comprising multiple isoforms that can be identified by their relative pharmacological profile towards PACAP38, PACAP27, and VIP. As shown in Fig. 1A, all three peptides stimulated aldosterone production with approximately the same efficacy (ED50 = 0.52 ± 0.10, 0.14 ± 0.023, 0.70 ± 0.22 nM, respectively) in human tu- morous H295R cells, indicating that PACAP effect was mediated by the type II PACAP/VIP receptor in these cells. Figure 1B shows that VIP had no stimulatory effect in bovine ZG cells, whereas PACAP38 and PACAP27 both stimulated aldosterone production with ED50 similar to those observed in human cells (0.33 ± 0.066 and 0.55 ± 0.096 nM, respectively). Thus, PACAP seemed to act through type I PACAP-specific receptor in bovine ZG cells. These results showed that PACAP could directly stimulate aldosterone-producing cells in normal bovine and human tumorous adrenal gland. However, they sug- gested a species-dependent and/or cell type-dependent mode of action, i.e. interaction with type I receptor in cow and type II receptor in human tumor cells.
Effects of PACAP and VIP on cAMP production in human and bovine adrenal cells
Both type I and type II PACAP receptor subtypes are positively coupled to AC. As shown in Fig. 2A, PACAP38, PACAP27, and VIP strongly stimulated cAMP production in a dose-dependent manner in human H295R cells with ED50 of 0.35 ± 0.020, 0.20 ± 0.014, and 0.38 ± 0.021 nM, respec- tively. In bovine ZG cells, only PACAP38 and PACAP27 increased cellular cAMP content, although to a lesser extent than in human H295R cells. The doses eliciting a half-max- imum increase in cAMP production were 0.12 ± 0.03 and 0.27 ± 0.042 nM, respectively (Fig. 2B). VIP at a concentration of 100 nm had no effect on AC activity in these cells. The ED50 values for cAMP production were in very good agreement with those reported for aldosterone production (Fig. 1). These results confirmed the absence of type II PACAP/VIP receptor in bovine ZG cells.
Effect of PACAP on inositol (1, 4, 5) trisphosphate production in human and bovine adrenal cells
Besides their coupling to AC, type I PACAP receptors also may activate PLC. Type II PACAP/VIP receptors are cou- pled only to AC. Treatment of human H295R cells with 100 nM PACAP38 or PACAP27 did not elicit any Ins(1, 4, 5)P3 production, whereas AII, which is well known to activate PLC in these cells, induced a 2.6-fold increase in intracellular Ins(1,4,5)P3 content (Fig. 3A). By contrast, in bovine ZG cells,
A. H295R cells
· PACAP38
ALDOSTERONE (fmol/well)
PACAP27
1000
VIP
800
1
600
400
200
-12
-11
-10
-9
-8
-7
LOG [PEPTIDE] (M)
B. ZG cells
· PACAP38
ALDOSTERONE (fmol/well)
PACAP27
1200.0
VIP
900.0
600.0
Ý
300.0
-12 -11 -10 -9 -8 -
-7
LOG [PEPTIDE] (M)
PACAP38 and PACAP27, at a concentration of 100 nm, elic- ited a statistically significant 2.1- to 2.3-fold increase in Ins(1, 4, 5)P3 production (Fig. 3B). These results further substanti- ated the differential expression of PACAP receptor type I and type II in bovine ZG and human H295R tumorous cells, respectively.
A. H295R cells
Intracellular cAMP (pmol/well)
PACAP38
PACAP27
5
T
₣
EE
VIP
4
3
3
2
1
-12
-11
-10
-9
-8
-7
LOG [PEPTIDE] (M)
B. ZG cells
Intracellular cAMP (pmol/well)
· PACAP38
1.2
PACAP27
3
VIP
1.0
5
0.8
0.6
0.4
0.2
-12
-11
-10
-9
-8
-7
LOG [PEPTIDE] (M)
Additive effects of PACAP and AII on human and bovine adrenal cells
AII lies among the most important physiological secreta- gogues of aldosterone. In all species investigated, including bovine and human, receptors responsible for this steroido-
A. H295R cells
*
Ins(1,4,5)P3 (pmol/well)
12
10
8
6
4
2
CTL
All
P38
P27
B. ZG cells
12
*
Ins(1,4,5)P3 (pmol/well)
*
10
*
8
6
4
T
2
CTL
All
P38
P27
genic effect are of the AT1 subtype, coupled to the activation of PLC-ß through a G protein of the Ga family. AII has no, or very slight, effect on cAMP production (18). When AII or PACAP38 were given alone to human H295R cells, they stimulated aldosterone production to approximately the same extent (29.31 ± 0.45 and 39.86 ± 1.12 fmol/well, re- spectively) (Fig. 4A). However, given together, their effects were additive, and aldosterone production was raised to 93.75 ± 4.50 fmol/well (P < 0.01). This result was in accor- dance with the proposal that AII and PACAP activated two distinct signaling pathways in human adrenocortical carci- noma H295R cells: Ins(1,4,5)P3 and DAG for AII, and cAMP for PACAP. The situation clearly was different in bovine ZG cells, where AII was a more efficacious secretagogue than PACAP38 (7.06 ± 0.76 and 4.38 ± 0.27 pmol/well, respec-
A.
H295R cells
*
ALDOSTERONE (fmol/well)
100
I
75
50
T
T
25
CTL
All
P38 All+P38
B. ZG cells
ALDOSTERONE (pmol/well)
8
T
6
T
4
2
T
CTL
All
P38 All+P38
tively), but most importantly, the effects of AII and PACAP were not additive [8.23 ± 0.27 pmol/well, not significant (NS)] in this species. This result suggested that PACAP38 and AII shared, at least in part, a common intracellular mecha- nism of action, i.e. the PLC-mediated pathway, in stimulating aldosterone production in bovine ZG cells.
Effect of ANP on PACAP-stimulated aldosterone production in human and bovine adrenal cells
The cardiac hormone ANP is the major physiological in- hibitor of aldosterone synthesis and secretion. Numerous studies on cultured ZG cells from various species (rat, bo- vine, human) have shown that ANP counteracts the effects of most known secretagogues of aldosterone (12). Recently we reported the presence of fully functional receptors of
ANP inhibiting AII-stimulated aldosterone production in the human adrenocortical H295R cell line (11). It was then tempt- ing to compare the ability of ANP to inhibit PACAP-stim- ulated aldosterone production in human tumorous and nor- mal bovine adrenal cells. ANP had no significant effect on the control level of aldosterone production (CTL: 19.71 ± 2.59 fmol/well, ANP 10 nM: 11.60 ± 0.53 fmol/well, NS). How- ever, simultaneous treatment of human H295R cells with PACAP38 or PACAP27 and ANP induced a modest, but yet statistically significant (P < 0.01), 25-27% inhibition of al- dosterone production (P38 10 nM: 182.25 ± 7.92 fmol/well, P38 10 nM + ANP 10 nM: 132.60 ± 5.81 fmol/well, P27 10 nM: 198.00 ± 7.81 fmol/well, P27 10 nM + ANP 10 nM: 148.50 ± 5.86 fmol/well). In bovine ZG cells, ANP effect was much more pronounced, with an 80-87% inhibition of PACAP- stimulated aldosterone production (CTL: 89.8 ± 8.0 fmol/ well, ANP 10 nM: 63.5 ± 6.6 fmol/well, NS; P38 10 nM: 1008.0 ± 118.8 fmol/well, P38 10 nM + ANP 10 nM: 197.5 ± 23.2 fmol/well, P27 10 nM: 1143.3 + 151.0 fmol/well, P27 10 nM + ANP 10 nM: 149.7 ± 19.8 fmol/well) This differential inhibitory effect of ANP on PACAP-stimulated aldosterone production in bovine and human adrenal cells could not be attributed to a difference in ANP potency in the two cellular models. Actually, the IC50 of ANP in inhibiting PACAP38- stimulated aldosterone production was 0.16 nm in human H295R cells and 0.12 nM in bovine ZG cells (data not shown). Rather, the difference in ANP efficacy in inhibiting aldoste- rone production in the two cell models could be attributed to a different mode of action of ANP in response to distinct signaling pathways elicited by PACAP in bovine and human adrenocortical cells. To test this hypothesis, we performed dose-response curves of PACAP38 in absence or presence of a maximally inhibiting concentration of ANP (10 nM). In human H295R tumorous cells, ANP caused a 35% inhibition of the maximal stimulatory effect of PACAP38 without af- fecting the ED50 of the curve (1.59 ± 0.26 nM in the absence and 1.06 ± 0.20 nm in the presence of ANP)(Fig. 5A). In contrast, in bovine ZG cells, ANP profoundly inhibited the maximal stimulatory effect of PACAP38 (74%) and also shifted the ED50 of the curve to higher concentration (0.52 ± 0.071 nM vs. 11.11 ± 6.46 nm) (Fig. 5B). The differential inhibitory effect of ANP on PACAP-stimulated aldosterone production in bovine ZG and human H295R cells could thus be the reflection of the different transduction pathways me- diating the effect of PACAP in the two cell models.
Discussion
Aldosterone secretion is known to be regulated in a para- crine manner by several neuropeptides released by the nerve plexuses distributed in the outer zone of the adrenal cortex (19). The present work demonstrates that the neuropeptide, PACAP, exerts a direct stimulatory effect on aldosterone production in bovine ZG and human adrenocortical carci- noma cells in culture. Previous studies in frog adrenal gland have shown that PACAP released from nerve endings may directly regulate aldosterone secretion through specific ad- renocortical receptors (6). However, because the adrenal tis- sue of amphibians is composed of a mixed population of steroid-producing cells and chromaffin cells, similar studies
The Endocrine Society. Downloaded from press.endocrine.org by [$ {individualUser.displayName} ] on 05 October 2015. at 14:33 For personal use only. No other uses without permission. . All rights reserved.
A. H295R cells
ALDOSTERONE (fmol/well)
· - ANP
0 + ANP
500
400
300
200
100
-12 -11 -10 -9 -8 -7
LOG [PACAP38] (M)
B. ZG cells
· - ANP
ALDOSTERONE (fmol/well)
150
0 + ANP
125
100
75
50
25
-12
-11
-10
-9
-8
-7
LOG [PACAP38] (M)
in mammalian species had to be performed. Nussdorfer and co-workers (7,8) have shown that PACAP seems to indirectly stimulate aldosterone production in rat and human adrenal glands by eliciting catecholamine release from neighboring
medullary chromaffin cells. The apparent discrepancy be- tween their study and the present work could be explained, at least in part, by differences in the methodology used: dispersed cells vs. cultured cells. In our hands, freshly dis- persed ZG cells are poorly responsive to any secretagogue because they are already in an activated state and secrete high levels of aldosterone immediately after the collagenase dispersion step. It is possible also that rat ZG cells might not express PACAP-specific receptors, as cytoautoradiographic studies in rat adrenal gland seem to demonstrate (5). Our study is therefore the first to report a direct steroidogenic action of PACAP in mammalian adrenal gland.
Comparison of the effect of PACAP38, PACAP27, and VIP on aldosterone production in the two cell system models we used suggests that PACAP interacts with a different receptor subtype in human adrenocortical carcinoma cells and bovine ZG cells (Fig. 1). A comparative study of the second mes- sengers generated by PACAP and VIP in the two cell systems further substantiates this hypothesis (Figs. 2 and 3). To- gether, the results suggest that the direct steroidogenic action of PACAP is mediated through PVR1 (type I PACAP-specific receptors) coupled to both PLC and AC in bovine ZG cells but through PVR2 or PVR3 (type II PACAP/VIP receptors) coupled to AC in human H295R cells.
The rat PVR1 exists in six variant forms, a short form (PVR1s) and five splice variants having inserts in the third intracellular loop of the receptor, a domain believed to be important for interaction with G proteins (2). There are two distinct 28-amino acid inserts (termed hip and hop1), a 27- amino acid insert (hop2), and two combination inserts (termed hip-hop1 and hip-hop2). The presence of the hop insert alone has no influence on the potency of cAMP and Ins(1, 4, 5)P3 production. The hip insert alone abolishes cou- pling to PLC and alters potency of coupling to AC. The combination of the two inserts, hip-hop, gives an interme- diate phenotype displaying slightly altered efficiency for AC stimulation and requiring high concentration (100 nm) for the stimulation of PLC. Likewise, four splice variants of the human PVR1 have been characterized, and the short and hop forms of the bovine PVR1 have been cloned (20, 21). In bovine ZG cells, PACAP stimulated cAMP production very potently (ED50 = 0.12-0.27 nm), but at a lower level than in human H295R cells, and was able to activate PLC. This suggests that PACAP receptors in these bovine adrenal cells are either of the PVR1s or the PVR1hop subtype if we refer to the rat nomenclature.
The effect of PACAP was compared with that of AII, the main physiological secretagogue of aldosterone. In human H295R cells, PACAP was as efficient as AII; both hormones induced a 3.6- to 4.9-fold increase in aldosterone production. In addition, their effects were additive in this species, with an 11.4-fold increase in steroid production with combined treatment. These results clearly suggest that PACAP and AII activate two distinct signaling pathways (cAMP and Ins(1, 4, 5)P3/Ca2+, respectively) in human adrenocortical carcinoma cells. The situation was strikingly different in bovine ZG cells, where AII was consistently more efficient than PACAP, with a 9.6-fold increase in aldosterone production as com- pared with a 5.9-fold increase induced by PACAP. Moreover, AII stimulation could not be enhanced by simultaneous treat-
ment with PACAP. This situation is reminiscent of the cross- talk between GnRH and PACAP action in gonadotropes (1). In these cells, PACAP acts through PVR1 and stimulates both AC and PLC, whereas GnRH acts through a Gq-coupled receptor and stimulates only PLC. In both normal and clonal gonadotropes, the synergistic effect of PACAP on GnRH- stimulated gonadotropin release is only observed at lower (10 pM) concentrations, where PACAP would be expected to preferentially stimulate the cAMP/PKA pathway (PACAP is more potent in stimulating AC than PLC in these cells). At maximally effective concentration of both factors, GnRH in- hibits PACAP-stimulated cAMP production, and synergy could no longer be observed because the two factors activate the same intracellular mechanism.
The last part of the present study concerns the effect of the natriuretic hormone ANP on PACAP-stimulated aldoste- rone secretion. As is true for all known secretagogues of aldosterone, ANP is able to inhibit the effect of PACAP in both human and bovine adrenocortical cells. However, the extent of this inhibition depends on the species and/or cell type concerned, being greater in cow than in human tumor cells. Barrett et al. (22) examined the inhibitory effect of ANP on aldosterone secretion stimulated by agonists that use ei- ther the Ca2+-phosphoinositide (AII) or the cAMP (ACTH) messenger system in bovine ZG cells. In these cells, ANP action is mediated by both a cyclic nucleotide-dependent and -independent pathway. Whereas the former is attributed to the activation of a cGS-PDE that hydrolyses the cAMP pro- duced by ACTH activation (23), the latter operates through a yet unknown mechanism presumably acting downstream from the step of PLC activation of inositol phosphates pro- duction and intracellular calcium mobilization. The modest inhibitory effect of ANP on PACAP-stimulated aldosterone production in human H295R cells, where the neuropeptide activates the cAMP pathway exclusively, could be tentatively explained by analogy to ANP inhibition of ACTH-stimulated aldosterone production in bovine and rat ZG cells as de- scribed above. In addition, the lack of effect of ANP on PACAP ED50 on aldosterone production suggests that its inhibitory effect on cAMP intracellular accumulation is not altering a putative PACAP receptor reserve typically asso- ciated with a hyperbolic coupling of stimulus with cellular response (24). This is confirmed by our observation that PACAP ED50 on cAMP accumulation and on aldosterone production are superimposable, suggesting a linear coupling between second-messenger production and cellular re- sponse. These results are in sharp contrast with those ob- tained in bovine ZG cells, where ANP not only drastically inhibited PACAP-stimulated aldosterone production but also shifted to the right the concentration-response curve for PACAP. In this case, where PACAP stimulates both Ca2+- phosphoinositide and cAMP pathways, the cyclic nucleoti- de-independent pathway of ANP action seems to prevail over the cyclic nucleotide-dependent pathway. The right- ward shift in the concentration-response curve produced by ANP could be explained in light of the model of functional antagonism developed by Van den Brink based on the con- cept of receptor reserve (24). This model predicts that, in a tissue or a cell type with a large receptor reserve, increasing the concentration of the antagonist will result initially in
parallel shifts in the concentration-response curve to the ag- onist. However, when the receptor reserve is abolished, fur- ther increase in antagonist concentration will cause a dimi- nution in the maximal response to the agonist. Because it is generally agreed that ANP does not alter inositol phosphates production or intracellular calcium mobilization (12), the nonlinear coupling step attenuated by ANP in bovine ZG cells, which also would be responsible for the apparent re- ceptor reserve displayed, is probably located downstream from second messenger production.
In conclusion, investigation of the cross-talk between PACAP and ANP signaling pathways in two different cell system models bearing a different PACAP receptor subtype, which activates different signaling pathways, could improve our understanding of the intracellular mechanism of action of PACAP and ANP in regulating aldosterone biosynthesis.
Acknowledgment
The authors wish to thank Normand McNicoll for his stimulating discussions and judicious advice.
References
1. Rawlings SR, Hezareh M 1996 Pituitary adenylate cyclase-activating polypep- tide (PACAP) and PACAP/vasoactive intestinal polypeptide receptors: ac- tions on the anterior pituitary gland. Endocr Rev 17:4-29
2. Journot L, Waeber C, Pantaloni C, Holsboer F, Seeburg PH, Bockaert J, Spengler D 1995 Differential signal transduction by six splice variants of the pituitary adenylate cyclase-activating peptide (PACAP) receptor. Biochem Soc Trans 23:133-137
3. Tabarin A, Chen D, Håkanson R, Sundler F 1994 Pituitary adenylate cyclase- activating peptide in the adrenal gland of mammals: distribution, character- ization and responses to drugs. Neuroendocrinology 59:113-119
4. Takahashi K, Totsune K, Murakami O, Sone M, Itoi K, Miura Y, Mouri T 1993 Pituitary adenylate cyclase-activating polypeptide (PACAP)-like immu- noreactivity in pheochromocytomas. Peptides 14:365-369
5. Shivers BD, Görcs TJ, Gottschall PE, Arimura A 1991 Two high affinity binding sites for pituitary adenylate cyclase-activating polypeptide have dif- ferent tissue distribution. Endocrinology 128:3055-3065
6. Yon L, Chartel n Feuilloley M, De Marchis S, Fournier A, De Rijk E, Pelletier G, Roubos E, Vaudry H 1994 Pituitary adenylate cyclase-activating polypep- tide stimulates both adrenocortical cells and chromaffin cells in the frog ad- renal gland. Endocrinology 135:2749-2758
7. Andreis PG, Malendowicz LK, Belloni AS, Nussdorfer GG 1995 Effects of pituitary adenylate-cyclase activating peptide (PACAP) on the rat adrenal secretory activity: preliminary in-vitro studies. Life Sci 56:135-142
8. Neri G, Andreis PG, Prayer-Galetti T, Rossi GP, Malendowicz LK, Nussdorfer GG 1996 Pituitary adenylate-cyclase activating peptide enhances aldosterone secretion of human adrenal gland: evidence for an indirect mech- anism, probably involving the local release of catecholamines. J Clin Endo- crinol Metab 81:169-173
9. Edwards AV, Jones CT 1994 Adrenal responses to the peptide PACAP in conscious functionally hypophysectomized calves. Am J Physiol 266:E870-E876
10. Bird IM, Hanley NA, Word AR, Mathis MJ, McCarthy JL, Mason IJ, Rainey WE 1993 Human NCI-H295 adrenocortical carcinoma cells: a model for an- giotensin II-responsive aldosterone secretion. Endocrinology 133:1555-1561
11. Bodart V, Rainey WE, Fournier A, Ong H, De Léan A 1996 The human H295R adrenocortical cell line contains functional atrial natriuretic peptide receptors that inhibit aldosterone biosynthesis. Mol Cell Endocrinol 118:137-144
12. Ganguly A 1992 Atrial natriuretic peptide-induced inhibition of aldosterone secretion: a quest for mediator(s). Am J Physiol 263:E181-E194
13. De Léan A, Racz K, McNicoll N, Desrosiers ML 1984 Direct ß-adrenergic stimulation of aldosterone secretion in cultured bovine adrenal subcapsular cells. Endocrinology 115:485-492
14. Brochu M, Féthière J, Roy M, Ong H, De Léan A 1989 Highly sensitive and rapid radioimmunoassay for aldosterone in plasma and cell culture medium. Clin Biochem 22:289-292
15. Brown BL, Albano JDM, Ekins RP, Sgherzi AM, Tampion W 1971 A simple and sensitive saturation assay method for the measurement of adenosine 3’:5’-cyclic monophosphate. Biochem J 121:561-562
16. Palmer S, Hughes KT, Lee DY, Wakelam MJO 1989 Development of a novel, Ins(1,4,5)P3-specific binding assay. Its use to determine the intracellular con-
centration of Ins(1,4,5)P3 in unstimulated and vasopressin-stimulated rat hepa- tocytes. Cell Signal 1:147-156
17. De Léan A, Munson PJ, Rodbard D 1978 Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay and physiological dose-response curves. Am J Physiol 235:E97-E102
18. Fujita K, Aguilera G, Catt KJ 1979 The role of cyclic AMP in aldosterone production by isolated zona glomerulosa cells. J Biol Chem 254:8567-8574
19. Holzwarth MA, Cunningham LA, Kleitman N 1987 The role of adrenal nerves in the regulation of adrenocortical functions. Ann NY Acad Sci 512:449-463
20. Pisegna JR, Wank SA 1996 Cloning and characterization of the signal trans- duction of four splice variants of the human pituitary adenylate cyclase- activating polypeptide receptor. J Biol Chem 271:17267-17274
21. Miyamoto Y, Habata Y, Ohtaki T, Masuda Y, Ogi K, Onda H, Fujino M 1994 Cloning and expression of a complementary DNA encoding the bovine re- ceptor for pituitary adenylate cyclase-activating polypeptide (PACAP). Bio- chim Biophys Acta 1218:297-307
22. Barrett PQ, Isales CM 1988 The role of cyclic nucleotides in atrial natriuretic peptide-mediated inhibition of aldosterone secretion. Endocrinology 122:799-808
23. MacFarland RT, Zelus BD, Beavo JA 1991 High concentrations of a cGMP- stimulated phosphodiesterase mediate ANP-induced decreases in cAMP and steroidogenesis in adrenal glomerulosa cells. J Biol Chem 266:136-142
24. Van den Brink FG 1973 The model of functional interaction. I. Development and first check of a new model of functional synergism and antagonism. Eur J Pharmacol 22:270-278