0013-7227/90/1262-0849$02.00/0 Endocrinology Copyright© 1990 by The Endocrine Society
Vol. 126, No. 2 Printed in U.S.A.
Roles of Intracellular and Extracellular Calcium in the Kinetic Profile of Adrenocorticotropin Secretion by Perifused Rat Anterior Pituitary Cells. I. Corticotropin Releasing Factor Stimulation* JUSTIN G. S. WONf AND DAVID N. ORTH Departments of Medicine, (J.G.S.W., D.N.O.) and Molecular Physiology and Biophysics (D.N.O.), Vanderbilt University Medical Center, Nashville, Tennessee 37232
respectively, and the responses to both agents were delayed by 1 min. Preperifusion of the cells with 5 pM penfluridol, a calmodulin inhibitor, reduced CRF- and 8-Br-cAMP-induced ACTH release by 54% and 41%, respectively. The combination of Ca2+i depletion and perifusion with 100 nM dexamethasone, a maximally inhibitory concentration, inhibited CRF- and 8-BrcAMP-stimulated ACTH release by 82% and 83%, respectively. These results indicate that 1) CRF and 8-Br-cAMP both require Ca2+e influx to elicit maximal ACTH release, consistent with the concept that CRF acts mainly via activation of the adenylate cyclase/cAMP-dependent protein kinase-A pathway, causing influx of Ca2+e; 2) CRF- and 8-Br-cAMP-stimulated entry of Ca2+e proceeds via both L- and non-L-type voltageregulated Ca2+ channels; 3) at least one site of Ca2+'s action in effecting ACTH release lies distal to cAMP generation; 4) Ca2+; may be involved in the initial incremental phase of the response to CRF that is analogous to the transient spike phase of the response to agents, such as arginine vasopressin, which act via the Ca2+/inositol phosphate-dependent protein kinase-C pathway; 5) both phases of the response to CRF appear to involve Ca2+-dependent calmodulin; 6) at least one site at which glucocorticoids exert their inhibitory effect lies distal to cAMP generation; 7) glucocorticoids appear to inhibit secretion of the Ca2+-regulated ACTH pool; and, finally, 8) there appears to be a fraction of CRF-stimulated ACTH secretion that is both Ca2+ independent and glucocorticoid nonsuppressible. (Endocrinology 126: 849-857, 1990)
ABSTRACT. We examined the effects of removing extracellular Ca2+ (Ca2+e), depleting intracellular Ca2+ (Ca2+i), inhibiting Ca2+-dependent calmodulin, blocking voltage-sensitive Ca2+ channels, and combining Ca2+i depletion with exposure to glucocorticoid on the secretion of ACTH by perifused dispersed rat anterior pituitary cells stimulated with ovine CRF and 8-bromocAMP (8-Br-cAMP). A time-course analysis of the effect of perifusing the cells for 60 min with Ca2+-free medium on 10 nM CRF-stimulated ACTH release revealed that inhibition required about 3 min to begin and about 40 min to reach maximal effect. Within 2 min of restoring Ca2+ to the medium, ACTH secretion rebounded for about 5 min before falling to the pre-Ca2+e removal rate. A similar pattern and time course were observed when Ca2+e was more completely removed by perifusing the cells with Ca2+-free medium containing 2 mM EGTA, except that greater suppression was observed. Removing Ca2+e reduced CRF- and 5 mM 8-Br-cAMP-induced ACTH release by 54% and 49%, respectively, and delayed by 1 min the response to 8-Br-cAMP, but not that to CRF. Perifusing 0.2 mM nimodipine, a dihydropyridine Ca2+ channel blocker, before and after restoration of Ca2+ to the Ca2+-free medium inhibited ACTH release by 40-48%, and the blockade persisted for at least 70 min after nimodipine was removed from the medium. When intracellular Ca2+ was depleted by perifusing the cells with Ca2+-free/EGTA medium containing the Ca2+ ionophore A23187 to facilitate the efflux of Ca2+i( CRF- and 8-BrcAMP-stimulated ACTH release were reduced by 70% and 71%,
Ca2+ channels (4-7), and the mechanism of action of the increased cytosolic free Ca2+; in the secretory process are much less well understood. Furthermore, most previous studies have examined secretion over intervals of hours, whereas endogenous ACTH secretion is pulsatile, with individual pulses lasting only a few minutes (8). In the present study we examined the effects of removing Ca2+e, depleting Ca2+i, and blocking Ca2+ channels on the minute by minute kinetic profile of CRFstimulated ACTH release from dispersed normal rat anterior pituitary cells in a microperifusion system (9). We have attempted to define further the step(s) at which Ca2+ acts in the protein kinase-A signal transduction
C
ALCIUM is known to play an important role in CRF-stimulated secretion of ACTH by the rat anterior pituitary gland (1-6). However, the relative importance of intra- and extracellular Ca2+ (Ca2+; and Ca2+e, respectively) (2-4), the roles of the different types of Received September 5,1989. Address all correspondence to: Dr. David N. Orth, Division of Endocrinology, AA4206 Vanderbilt Medical Center North, Nashville, Tennessee 37232. * This work was supported in part by Research Grants 5-R01-DK33334 and 5-M01-RR-00095 from the NIH, USPHS. t Current address: Section of Endocrinology and Metabolism, Veterans General Hospital, Shih-pai, Taipei, Taiwan 11217, Republic of China.
849
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
ROLE OF Ca2+ IN CRF-STIMULATED ACTH RELEASE
850
pathway and whether glucocorticoids regulate Ca2+-dependent or -independent ACTH secretion.
Materials and Methods Primary cell cultures Male Sprague-Dawley rats, weighing 150-200 g, were killed by decapitation. Anterior lobes of the pituitary glands were rapidly removed, and the cells were enzymatically dispersed as previously described (10) with minor modifications. Briefly, the lobes were rinsed with HEPES buffer [137 mM NaCl, 5 mM KC1, 0.7 mM Na2HPO4, and 25 mM HEPES (Sigma Chemical Co., St. Louis, MO), pH 7.3, containing 10 mM glucose and 0.4% (wt/vol) BSA (crystalline BSA, Sigma)], finely minced, and dissociated with HEPES buffer containing 0.4% collagenase (type II, Worthington Biochemical Corp., Freehold, NJ) and 40 Mg/ml DNase (type I, Worthington) in a sterile 25-ml Spinner flask (Bellco Glass Co., Vineland, NJ) partially immersed in a 37 C water bath, and stirred at 100-200 rpm for 90 min. The cells were rinsed three times by centrifugation and resuspension in HEPES buffer, mechanically dispersed by repeated gentle aspiration and expulsion through a 1-ml plastic pipette tip (Rainin Instrument Co., Woburn, MA), filtered through nylon mesh (44 jum; Small Parts, Inc., Miami, FL), and centrifuged through HEPES buffer containing a cushion of 20% BSA to remove the majority of erythrocytes (11). The supernatant medium was discarded, and the pelleted cells were resuspended in tissue culture medium [50% Dulbecco's Modified Eagle's Medium (Sigma) and 50% Ham's F-12 medium (Gibco, Grand Island, NY), containing 25 mM HEPES, 29 mM NaHCO3, 200 U/ml penicillin* 200 Mg/ml streptomycin, 50 ng/ ml gentamicin (all from Sigma), and 15% heat-inactivated fetal calf serum (Gibco); 5 jug/ml insulin (Sigma), 5 Mg/ml transferrin (Sigma), and 1 ng/ml fibroblast growth factor (Collaborative Research, Bedford, MA) were added immediately before use. The yield was 1.5-2 x 106 cells/anterior pituitary. Cell viability, as assessed by exclusion of trypan blue, was greater than 90%. The dissociated cells were carefully layered on preswollen Sephadex G-10 resin (Pharmacia Fine Chemicals, Piscataway, NJ) in 48-well cluster dishes (Costar Corp., Cambridge, MA; 0.50.7 x 106 cells/well) and cultured for 3-5 days in a humidified CO2 incubator as previously described (12). The cells and G-10 resin from a single well were agitated by gentle aspiration and expulsion through a 200-^1 Pipetteman (Rainin Instrument Co., Woburn, MA) plastic tip and were loaded into a single microperifusion column. Reagents Synthetic ovine CRF was generously provided by Dr. Jean Rivier of the Salk Institute (La Jolla, CA), anticorticotropin serum IgG-ACTH-1 was a gift of IgG Corp. (Nashville, TN), nimodipine a gift of Miles Pharmaceuticals (West Haven, CT), and penfluridol a gift of Janssen Research Foundation (Beerse, Belgium). Calcium ionophore A23187, 8-bromo-cAMP (8-BrcAMP), EGTA, and dexamethasone were purchased from Sigma. Penfluridol and dexamethasone were dissolved in ethanol, and nimodipine and ionophore A23187 were dissolved in dimethylsulfoxide, after which they all were diluted to their
Endo • 1990 Voll26«No2
final concentrations with perifusion medium. The final concentration of ethanol or dimethylsulfoxide in the perifusion medium did not exceed 0.1%. Microperifusion system The microperifusion system and its operation have been described previously (9). The perifusion medium was Eagle's Minimum Essential Medium [constituted from Minimum Essential Medium amino acids (Gibco 320-1135 AG), vitamins (Gibco 320-1120 AG), 200 mg/liter CaCl2 (anhydrous), 400 mg/ liter KC1, 97.67 mg/liter MgSO4 (anhyd.), 6.8 g/liter NaCl, 2.2 g/liter NaHCO3, 1.4 g/liter NaH 2 PO 4 H 2 O, 1 g/liter D-glucose, 6 mg/liter phenol red, 100 mg/liter sodium succinate, and 75 mg/liter succinic acid], supplemented with 10 mM HEPES, 0.1% (wt/vol) BSA, 100 U/ml penicillin, and 100 Mg/ml streptomycin. For Ca2+-free incubations, the perifusion medium consisted of Eagle's Minimal Essential Medium (Gibco) lacking CaCl2 (1.8 mM Ca2+) and supplemented with the same additives; Mg2"1" was not substituted for the missing Ca2+. One-minute (72-/ul) effluent fractions were collected in tubes containing 728 n\ RIA diluent (13) and stored at -20 C until immediately before RIA. RIA The ACTH RIA was performed using anticorticotropin serum IgG-ACTH-1, as previously described (13). Statistical analysis Results are expressed as the mean ± SEM. Statistical analyses were performed by one-way analysis of variance (ANOVA), followed by Duncan's multiple range test or Student's paired t test, as appropriate; P < 0.05 was considered significant. Calculations were performed with the CLINFO computer system of the Vanderbilt University General Clinical Research Center. The rebound phase of ACTH secretion in response to Ca2+-free or Ca2+-free/EGTA medium and the rates of decline in ACTH secretion after Ca2+ was removed us. after CRF was discontinued were analyzed by the multivariate ANOVA (MANOVA) program of SPSS-X, followed by Student's paired t test.
Results Time course of effect of Ca2+ removal or depletion and subsequent restoration
When the cells were perifused with complete medium, the ACTH concentration in effluent fractions reached a plateau within 3 min of exposure of the cells to CRF and remained relatively constant thereafter. Within 3 min after Ca2+-free medium was substituted, ACTH concentrations began to fall, fell rapidly for about 4 min, and then declined more gradually to reach a nadir 60% lower than the CRF-stimulated secretory rate after about 40 min, remaining relatively stable thereafter (Fig. 1A). Within 2 min of the time that complete medium was restored, ACTH secretion rebounded rapidly to reach
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
ROLE OF Ca2+ IN CRF-STIMULATED ACTH RELEASE 140 -
120 100 80 60 40 20 0 0
20
TIME AFTER Ca
40 2+
60
REMOVAL
80 -
100
min
120 100 -
Effect of removing Ca2+e on responses to CRF and 8-BrcAMP
5030 -
10
J
10
20
15
TIME AFTER REMOVAL - min
FIG. 1. Effect of removal of Ca 2+ 0 on the ACTH secretory response to
CRF. Four identical chambers were each loaded with 0.7 X 106 4-day cultured cells and perfused sequentially with complete perifusion medium containing 10 nM CRF for 20 min, Ca2+-free medium containing 10 nM CRF for 60 min, complete medium containing 10 nM CRF for 20 min, and complete medium alone for another 20 min. A, Immunoreactive ACTH (IR-ACTH) concentrations in 1-min (72-^1) effluent fractions collected in tubes containing 328 /xl RIA diluent. The boxes indicate the periods during which the cells were exposed to synthetic ovine CRF and Ca2+-free medium. Each point represents the mean of four determinations; the brackets indicate the SEM. *, P < 0.05; **, P < 0.01 (by Student's paired t test, vs. mean CRF-stimulated ACTH secretion during -18 through 0 min, before Ca2+ removal). B, IRACTH concentrations in fractions after Ca2+ removal (•; 1-20 min of A) and after cessation of CRF perifusion (•; 81-100 min of A) plotted on a semilogarithmic scale.
When tested 50 min after beginning perifusion with Ca2+-free medium, the integrated response to CRF was decreased by 54% (738 ± 58 us. 1619 ± 80 pg/20 min; P < 0.001), and the response to 8-Br-cAMP, which bypasses the activation of adenylate cyclase and directly stimulates protein kinase-A, was decreased by 49% (737 ± 52 vs. 1451 ± 48 pg/20 min; P < 0.001; Fig. 3). The increase in ACTH secretion after 8-Br-cAMP was delayed by 1 min with respect to the response to CRF (Fig. 3); this presumably reflected the additional time required
for 8-Br-cAMP to diffuse across the cell membrane. The 120 -i 'tract ion
r
EGTA medium was similar, but the inhibition of secretion was greater (87%), and the rebound after restoration of Ca2+ was also greater (P < 0.001, by MANOVA; Fig. 2). Because it required about 40 min for the maximum effect of Ca2+e removal to be achieved, cells were preperifused with the appropriate Ca2+-deficient medium for 50 min before they were challenged with other agents. In the remaining experiments, the calcium ionophore A23187 was added to the Ca2+-free/EGTA medium to facilitate the efflux of Ca2+; and thereby deplete the cells of Ca2+i. In experiments in which we tested the responses to CRF after removal of Ca2+e by 50-min preperifusion with Ca2+-free/EGTA medium or depletion of Ca2+; by 50-min preperifusion with Ca2+-free/EGTA/A23187 medium, there was no significant difference between the results (data not shown).
100 80
en a.
1
60
VCTH
-20
851
40
•**.
1
cc
20 0
2+
levels significantly higher than before Ca removal. Then, over a period of about 5 min, they fell to the previous plateau. After CRF was stopped, ACTH secretion fell, rapidly at first and then more gradually, to reach baseline within 20 min. The rate of decline during the first 20 min after CRF was discontinued was more rapid (P < 0.001, by MANOVA) than after Ca2+ was removed from the medium (Fig. IB). The kinetic profile of the effect of more complete removal of Ca2+e by perifusing the cells with Ca2+-free/
J
-20
20
40
60
TIME AFTER Ca 2+ REMOVAL
80 -
100
min
Ca2+e
FIG. 2. Effect of more complete removal of on the ACTH secretory response to CRF. Four identical chambers were each loaded with 0.7 x 106 4-day cultured cells and perfused sequentially with complete perifusion medium containing 10 nM CRF for 20 min, Ca2+-free medium containing 2 mM EGTA and 10 nM CRF for 60 min, complete medium containing 10 nM CRF for 20 min, and complete medium alone for another 20 min. One-minute (72-/ul) effluent fractions were collected in tubes containing 328 /nl RIA diluent. Data are plotted as described in Fig. 1A. IR-ACTH, Immunoreactive ACTH.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
ROLE OF Ca2+ IN CRF-STIMULATED ACTH RELEASE
Q.
I X h-
o I DC
140 - 8-Br-cAMP
120 -
120 -
100 H
100 -
80 -
80
60 -
60
40 -
40
20 -
20
0-
010
20
10
20
TIME AFTER STIMULATION - min FIG. 3. Effect of removal of Ca2+e on the ACTH secretory responses to CRF and 8-Br-cAMP. In both experiments, which were performed on different days, six identical chambers were each loaded with 0.7 x 106 4-day cultured cells. Cells in two sets of three randomly assigned chambers were perifused sequentially with either complete medium or Ca2+-free medium for 50 min, the same medium containing either 10 nM CRF or 5 mM 8-Br-cAMP for 20 min, and complete medium alone for another 20 min. The chambers that had initially been perfused with complete medium were then perfused with Ca2+-free medium for 50 min and finally with Ca2+-free medium containing either CRF or 8-BrcAMP for 20 min. The chambers that had initially been perfused with Ca2+-free medium were perfused with complete medium for 50 min and finally with complete medium containing either CRF or 8-Br-cAMP for 20 min. One of the six chambers in the 8-Br-cAMP experiment became occluded; data from the remaining five chambers were analyzed. IR-ACTH, Immunoreactive ACTH. • , Secretion by cells perifused with complete medium; O, secretion by cells perifused with Ca2+-free medium. Each point represents the mean of six (for CRF) or five (for 8-Br-cAMP) determinations; brackets indicate the SEM. *, P < 0.05 [by Student's paired t test, vs. basal secretion (mean IR-ACTH content of the five fractions collected during - 4 through 0 min)].
rate of increase in ACTH secretion during the initial incremental phase of the response to CRF (9, 14) was identical in complete and in Ca2+-free medium, but the initial rate of increase after stimulation with 8-Br-cAMP was slower in Ca2+-free medium (Fig. 3). Unless removing Ca2+ from the medium slows the diffusion of 8-Br-cAMP across the membrane, this difference suggests that CRF may act via a mechanism other than activation of cAMPdependent protein kinase-A to evoke the initial incremental ACTH secretory response. Effect of Ca2+ channel blockade on the response to CRF
Others have reported that Ca2+ channel blockers, including nimodipine, reduce the ACTH secretory response to CRF (5-7). We duplicated these results in monolayer cultures of rat anterior pituitary cells (data not shown), but were surprised to find that addition of up to 0.2 mM nimodipine to complete medium for periods as long as 1 h had no effect on the response to CRF in the microperifusion system. We considered that light might be inactivating the agent, so the entire apparatus, excluding the
Endo • 1990 Voll26«No2
fraction collector, was covered with aluminum foil. There was still no effect (data not shown). Therefore, the following experiment was performed to expose the cells to nimodipine before they were perifused with complete Ca2+-containing medium. Cells in four identical perifusion chambers were exposed simultaneously to four 20min pulses of CRF at intervals of 50 min. The first and last CRF pulses were applied after the cells had been perifused with complete medium for 50 min. The second CRF pulse was applied after the cells had been perifused with Ca2+-free medium for 50 min, and the third pulse was delivered after a 20-min perifusion with Ca2+-free medium containing nimodipine, followed by a 30-min perifusion with complete medium containing the same 0.2-mM concentration of nimodipine. As in previous experiments, removal of Ca2+e reduced the ACTH secretory response to CRF by 74% (303 ± 21 vs. 1168 ± 27 pg/20 min; P < 0.001; Fig. 4). Addition of nimodipine to complete medium inhibited the response by 40% (698 ± 99 vs. 1168 ± 27 pg/20 min; P < 0.001). Furthermore, the response to CRF 50 min after removal of nimodipine from the medium was reduced by 27% (855 ± 90 vs. 1168 ± 27 pg/20 min; P < 0.001), suggesting that nimodipine had a prolonged effect after it had been removed from the perifusion medium. To examine further the effects of nimodipine, six identical perifusion chambers were randomly assigned to one of two groups. The first two CRF pulses were applied to both groups under the same conditions as previously. For the third pulse, however, the chambers in one group were perfused with nimodipine in Ca2+-free medium for 20 min and then with nimodipine in complete medium for 30 min, whereas the other group was perfused with Ca2+-FREE
120 -
action
.2 o 2
140 -i CRF
NIMODIPINE
100 80 -
Q.
I X i—
R-AC
852
60 40 20 0
J
0
10
20
70
80
90
110 140 150 160 210 220 230
TIME - min
FIG. 4. Effect of nimodipine, a Ca2+ channel blocker, on the ACTH secretory response to CRF. Four identical chambers were each loaded with 0.7 x 106 4-day cultured cells and were perfused with 10 nM CRF for 20 min at 50-min intervals on four occasions: in complete medium, in Ca2+-free medium, in Ca2+-free medium containing 0.2 mM nimodipine, and again in complete medium. The boxes indicate the periods during which the cells were exposed to CRF, Ca2+-free medium, and nimodipine. IR-ACTH, Immunoreactive ACTH. Each point represents the mean of four determinations; brackets indicate the SEM.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
Effect of depleting Ca2+i on responses to CRF and 8-BrcAMP Depletion of Ca2+; reduced the integrated ACTH response to CRF by 70% (648 ± 34 vs. 2147 ± 245 pg/20 min; P < 0.01) and the response to 8-Br-cAMP by 71% (504 ± 5 vs. 1757 ± 45 pg/20 min; P < 0.001; Fig. 6). As when Ca2+ was removed from the medium, the rate of increase in ACTH secretion was slower after 8-Br-cAMP in Ca2+-depleted cells. However, in contrast to when Ca2+ 1 ||
140 -i .2 0
Ca»-FREE~Ti NIMODIPINE |
12
Vol. 126, No. 2 Printed in U.S.A.
Roles of Intracellular and Extracellular Calcium in the Kinetic Profile of Adrenocorticotropin Secretion by Perifused Rat Anterior Pituitary Cells. I. Corticotropin Releasing Factor Stimulation* JUSTIN G. S. WONf AND DAVID N. ORTH Departments of Medicine, (J.G.S.W., D.N.O.) and Molecular Physiology and Biophysics (D.N.O.), Vanderbilt University Medical Center, Nashville, Tennessee 37232
respectively, and the responses to both agents were delayed by 1 min. Preperifusion of the cells with 5 pM penfluridol, a calmodulin inhibitor, reduced CRF- and 8-Br-cAMP-induced ACTH release by 54% and 41%, respectively. The combination of Ca2+i depletion and perifusion with 100 nM dexamethasone, a maximally inhibitory concentration, inhibited CRF- and 8-BrcAMP-stimulated ACTH release by 82% and 83%, respectively. These results indicate that 1) CRF and 8-Br-cAMP both require Ca2+e influx to elicit maximal ACTH release, consistent with the concept that CRF acts mainly via activation of the adenylate cyclase/cAMP-dependent protein kinase-A pathway, causing influx of Ca2+e; 2) CRF- and 8-Br-cAMP-stimulated entry of Ca2+e proceeds via both L- and non-L-type voltageregulated Ca2+ channels; 3) at least one site of Ca2+'s action in effecting ACTH release lies distal to cAMP generation; 4) Ca2+; may be involved in the initial incremental phase of the response to CRF that is analogous to the transient spike phase of the response to agents, such as arginine vasopressin, which act via the Ca2+/inositol phosphate-dependent protein kinase-C pathway; 5) both phases of the response to CRF appear to involve Ca2+-dependent calmodulin; 6) at least one site at which glucocorticoids exert their inhibitory effect lies distal to cAMP generation; 7) glucocorticoids appear to inhibit secretion of the Ca2+-regulated ACTH pool; and, finally, 8) there appears to be a fraction of CRF-stimulated ACTH secretion that is both Ca2+ independent and glucocorticoid nonsuppressible. (Endocrinology 126: 849-857, 1990)
ABSTRACT. We examined the effects of removing extracellular Ca2+ (Ca2+e), depleting intracellular Ca2+ (Ca2+i), inhibiting Ca2+-dependent calmodulin, blocking voltage-sensitive Ca2+ channels, and combining Ca2+i depletion with exposure to glucocorticoid on the secretion of ACTH by perifused dispersed rat anterior pituitary cells stimulated with ovine CRF and 8-bromocAMP (8-Br-cAMP). A time-course analysis of the effect of perifusing the cells for 60 min with Ca2+-free medium on 10 nM CRF-stimulated ACTH release revealed that inhibition required about 3 min to begin and about 40 min to reach maximal effect. Within 2 min of restoring Ca2+ to the medium, ACTH secretion rebounded for about 5 min before falling to the pre-Ca2+e removal rate. A similar pattern and time course were observed when Ca2+e was more completely removed by perifusing the cells with Ca2+-free medium containing 2 mM EGTA, except that greater suppression was observed. Removing Ca2+e reduced CRF- and 5 mM 8-Br-cAMP-induced ACTH release by 54% and 49%, respectively, and delayed by 1 min the response to 8-Br-cAMP, but not that to CRF. Perifusing 0.2 mM nimodipine, a dihydropyridine Ca2+ channel blocker, before and after restoration of Ca2+ to the Ca2+-free medium inhibited ACTH release by 40-48%, and the blockade persisted for at least 70 min after nimodipine was removed from the medium. When intracellular Ca2+ was depleted by perifusing the cells with Ca2+-free/EGTA medium containing the Ca2+ ionophore A23187 to facilitate the efflux of Ca2+i( CRF- and 8-BrcAMP-stimulated ACTH release were reduced by 70% and 71%,
Ca2+ channels (4-7), and the mechanism of action of the increased cytosolic free Ca2+; in the secretory process are much less well understood. Furthermore, most previous studies have examined secretion over intervals of hours, whereas endogenous ACTH secretion is pulsatile, with individual pulses lasting only a few minutes (8). In the present study we examined the effects of removing Ca2+e, depleting Ca2+i, and blocking Ca2+ channels on the minute by minute kinetic profile of CRFstimulated ACTH release from dispersed normal rat anterior pituitary cells in a microperifusion system (9). We have attempted to define further the step(s) at which Ca2+ acts in the protein kinase-A signal transduction
C
ALCIUM is known to play an important role in CRF-stimulated secretion of ACTH by the rat anterior pituitary gland (1-6). However, the relative importance of intra- and extracellular Ca2+ (Ca2+; and Ca2+e, respectively) (2-4), the roles of the different types of Received September 5,1989. Address all correspondence to: Dr. David N. Orth, Division of Endocrinology, AA4206 Vanderbilt Medical Center North, Nashville, Tennessee 37232. * This work was supported in part by Research Grants 5-R01-DK33334 and 5-M01-RR-00095 from the NIH, USPHS. t Current address: Section of Endocrinology and Metabolism, Veterans General Hospital, Shih-pai, Taipei, Taiwan 11217, Republic of China.
849
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
ROLE OF Ca2+ IN CRF-STIMULATED ACTH RELEASE
850
pathway and whether glucocorticoids regulate Ca2+-dependent or -independent ACTH secretion.
Materials and Methods Primary cell cultures Male Sprague-Dawley rats, weighing 150-200 g, were killed by decapitation. Anterior lobes of the pituitary glands were rapidly removed, and the cells were enzymatically dispersed as previously described (10) with minor modifications. Briefly, the lobes were rinsed with HEPES buffer [137 mM NaCl, 5 mM KC1, 0.7 mM Na2HPO4, and 25 mM HEPES (Sigma Chemical Co., St. Louis, MO), pH 7.3, containing 10 mM glucose and 0.4% (wt/vol) BSA (crystalline BSA, Sigma)], finely minced, and dissociated with HEPES buffer containing 0.4% collagenase (type II, Worthington Biochemical Corp., Freehold, NJ) and 40 Mg/ml DNase (type I, Worthington) in a sterile 25-ml Spinner flask (Bellco Glass Co., Vineland, NJ) partially immersed in a 37 C water bath, and stirred at 100-200 rpm for 90 min. The cells were rinsed three times by centrifugation and resuspension in HEPES buffer, mechanically dispersed by repeated gentle aspiration and expulsion through a 1-ml plastic pipette tip (Rainin Instrument Co., Woburn, MA), filtered through nylon mesh (44 jum; Small Parts, Inc., Miami, FL), and centrifuged through HEPES buffer containing a cushion of 20% BSA to remove the majority of erythrocytes (11). The supernatant medium was discarded, and the pelleted cells were resuspended in tissue culture medium [50% Dulbecco's Modified Eagle's Medium (Sigma) and 50% Ham's F-12 medium (Gibco, Grand Island, NY), containing 25 mM HEPES, 29 mM NaHCO3, 200 U/ml penicillin* 200 Mg/ml streptomycin, 50 ng/ ml gentamicin (all from Sigma), and 15% heat-inactivated fetal calf serum (Gibco); 5 jug/ml insulin (Sigma), 5 Mg/ml transferrin (Sigma), and 1 ng/ml fibroblast growth factor (Collaborative Research, Bedford, MA) were added immediately before use. The yield was 1.5-2 x 106 cells/anterior pituitary. Cell viability, as assessed by exclusion of trypan blue, was greater than 90%. The dissociated cells were carefully layered on preswollen Sephadex G-10 resin (Pharmacia Fine Chemicals, Piscataway, NJ) in 48-well cluster dishes (Costar Corp., Cambridge, MA; 0.50.7 x 106 cells/well) and cultured for 3-5 days in a humidified CO2 incubator as previously described (12). The cells and G-10 resin from a single well were agitated by gentle aspiration and expulsion through a 200-^1 Pipetteman (Rainin Instrument Co., Woburn, MA) plastic tip and were loaded into a single microperifusion column. Reagents Synthetic ovine CRF was generously provided by Dr. Jean Rivier of the Salk Institute (La Jolla, CA), anticorticotropin serum IgG-ACTH-1 was a gift of IgG Corp. (Nashville, TN), nimodipine a gift of Miles Pharmaceuticals (West Haven, CT), and penfluridol a gift of Janssen Research Foundation (Beerse, Belgium). Calcium ionophore A23187, 8-bromo-cAMP (8-BrcAMP), EGTA, and dexamethasone were purchased from Sigma. Penfluridol and dexamethasone were dissolved in ethanol, and nimodipine and ionophore A23187 were dissolved in dimethylsulfoxide, after which they all were diluted to their
Endo • 1990 Voll26«No2
final concentrations with perifusion medium. The final concentration of ethanol or dimethylsulfoxide in the perifusion medium did not exceed 0.1%. Microperifusion system The microperifusion system and its operation have been described previously (9). The perifusion medium was Eagle's Minimum Essential Medium [constituted from Minimum Essential Medium amino acids (Gibco 320-1135 AG), vitamins (Gibco 320-1120 AG), 200 mg/liter CaCl2 (anhydrous), 400 mg/ liter KC1, 97.67 mg/liter MgSO4 (anhyd.), 6.8 g/liter NaCl, 2.2 g/liter NaHCO3, 1.4 g/liter NaH 2 PO 4 H 2 O, 1 g/liter D-glucose, 6 mg/liter phenol red, 100 mg/liter sodium succinate, and 75 mg/liter succinic acid], supplemented with 10 mM HEPES, 0.1% (wt/vol) BSA, 100 U/ml penicillin, and 100 Mg/ml streptomycin. For Ca2+-free incubations, the perifusion medium consisted of Eagle's Minimal Essential Medium (Gibco) lacking CaCl2 (1.8 mM Ca2+) and supplemented with the same additives; Mg2"1" was not substituted for the missing Ca2+. One-minute (72-/ul) effluent fractions were collected in tubes containing 728 n\ RIA diluent (13) and stored at -20 C until immediately before RIA. RIA The ACTH RIA was performed using anticorticotropin serum IgG-ACTH-1, as previously described (13). Statistical analysis Results are expressed as the mean ± SEM. Statistical analyses were performed by one-way analysis of variance (ANOVA), followed by Duncan's multiple range test or Student's paired t test, as appropriate; P < 0.05 was considered significant. Calculations were performed with the CLINFO computer system of the Vanderbilt University General Clinical Research Center. The rebound phase of ACTH secretion in response to Ca2+-free or Ca2+-free/EGTA medium and the rates of decline in ACTH secretion after Ca2+ was removed us. after CRF was discontinued were analyzed by the multivariate ANOVA (MANOVA) program of SPSS-X, followed by Student's paired t test.
Results Time course of effect of Ca2+ removal or depletion and subsequent restoration
When the cells were perifused with complete medium, the ACTH concentration in effluent fractions reached a plateau within 3 min of exposure of the cells to CRF and remained relatively constant thereafter. Within 3 min after Ca2+-free medium was substituted, ACTH concentrations began to fall, fell rapidly for about 4 min, and then declined more gradually to reach a nadir 60% lower than the CRF-stimulated secretory rate after about 40 min, remaining relatively stable thereafter (Fig. 1A). Within 2 min of the time that complete medium was restored, ACTH secretion rebounded rapidly to reach
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
ROLE OF Ca2+ IN CRF-STIMULATED ACTH RELEASE 140 -
120 100 80 60 40 20 0 0
20
TIME AFTER Ca
40 2+
60
REMOVAL
80 -
100
min
120 100 -
Effect of removing Ca2+e on responses to CRF and 8-BrcAMP
5030 -
10
J
10
20
15
TIME AFTER REMOVAL - min
FIG. 1. Effect of removal of Ca 2+ 0 on the ACTH secretory response to
CRF. Four identical chambers were each loaded with 0.7 X 106 4-day cultured cells and perfused sequentially with complete perifusion medium containing 10 nM CRF for 20 min, Ca2+-free medium containing 10 nM CRF for 60 min, complete medium containing 10 nM CRF for 20 min, and complete medium alone for another 20 min. A, Immunoreactive ACTH (IR-ACTH) concentrations in 1-min (72-^1) effluent fractions collected in tubes containing 328 /xl RIA diluent. The boxes indicate the periods during which the cells were exposed to synthetic ovine CRF and Ca2+-free medium. Each point represents the mean of four determinations; the brackets indicate the SEM. *, P < 0.05; **, P < 0.01 (by Student's paired t test, vs. mean CRF-stimulated ACTH secretion during -18 through 0 min, before Ca2+ removal). B, IRACTH concentrations in fractions after Ca2+ removal (•; 1-20 min of A) and after cessation of CRF perifusion (•; 81-100 min of A) plotted on a semilogarithmic scale.
When tested 50 min after beginning perifusion with Ca2+-free medium, the integrated response to CRF was decreased by 54% (738 ± 58 us. 1619 ± 80 pg/20 min; P < 0.001), and the response to 8-Br-cAMP, which bypasses the activation of adenylate cyclase and directly stimulates protein kinase-A, was decreased by 49% (737 ± 52 vs. 1451 ± 48 pg/20 min; P < 0.001; Fig. 3). The increase in ACTH secretion after 8-Br-cAMP was delayed by 1 min with respect to the response to CRF (Fig. 3); this presumably reflected the additional time required
for 8-Br-cAMP to diffuse across the cell membrane. The 120 -i 'tract ion
r
EGTA medium was similar, but the inhibition of secretion was greater (87%), and the rebound after restoration of Ca2+ was also greater (P < 0.001, by MANOVA; Fig. 2). Because it required about 40 min for the maximum effect of Ca2+e removal to be achieved, cells were preperifused with the appropriate Ca2+-deficient medium for 50 min before they were challenged with other agents. In the remaining experiments, the calcium ionophore A23187 was added to the Ca2+-free/EGTA medium to facilitate the efflux of Ca2+; and thereby deplete the cells of Ca2+i. In experiments in which we tested the responses to CRF after removal of Ca2+e by 50-min preperifusion with Ca2+-free/EGTA medium or depletion of Ca2+; by 50-min preperifusion with Ca2+-free/EGTA/A23187 medium, there was no significant difference between the results (data not shown).
100 80
en a.
1
60
VCTH
-20
851
40
•**.
1
cc
20 0
2+
levels significantly higher than before Ca removal. Then, over a period of about 5 min, they fell to the previous plateau. After CRF was stopped, ACTH secretion fell, rapidly at first and then more gradually, to reach baseline within 20 min. The rate of decline during the first 20 min after CRF was discontinued was more rapid (P < 0.001, by MANOVA) than after Ca2+ was removed from the medium (Fig. IB). The kinetic profile of the effect of more complete removal of Ca2+e by perifusing the cells with Ca2+-free/
J
-20
20
40
60
TIME AFTER Ca 2+ REMOVAL
80 -
100
min
Ca2+e
FIG. 2. Effect of more complete removal of on the ACTH secretory response to CRF. Four identical chambers were each loaded with 0.7 x 106 4-day cultured cells and perfused sequentially with complete perifusion medium containing 10 nM CRF for 20 min, Ca2+-free medium containing 2 mM EGTA and 10 nM CRF for 60 min, complete medium containing 10 nM CRF for 20 min, and complete medium alone for another 20 min. One-minute (72-/ul) effluent fractions were collected in tubes containing 328 /nl RIA diluent. Data are plotted as described in Fig. 1A. IR-ACTH, Immunoreactive ACTH.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
ROLE OF Ca2+ IN CRF-STIMULATED ACTH RELEASE
Q.
I X h-
o I DC
140 - 8-Br-cAMP
120 -
120 -
100 H
100 -
80 -
80
60 -
60
40 -
40
20 -
20
0-
010
20
10
20
TIME AFTER STIMULATION - min FIG. 3. Effect of removal of Ca2+e on the ACTH secretory responses to CRF and 8-Br-cAMP. In both experiments, which were performed on different days, six identical chambers were each loaded with 0.7 x 106 4-day cultured cells. Cells in two sets of three randomly assigned chambers were perifused sequentially with either complete medium or Ca2+-free medium for 50 min, the same medium containing either 10 nM CRF or 5 mM 8-Br-cAMP for 20 min, and complete medium alone for another 20 min. The chambers that had initially been perfused with complete medium were then perfused with Ca2+-free medium for 50 min and finally with Ca2+-free medium containing either CRF or 8-BrcAMP for 20 min. The chambers that had initially been perfused with Ca2+-free medium were perfused with complete medium for 50 min and finally with complete medium containing either CRF or 8-Br-cAMP for 20 min. One of the six chambers in the 8-Br-cAMP experiment became occluded; data from the remaining five chambers were analyzed. IR-ACTH, Immunoreactive ACTH. • , Secretion by cells perifused with complete medium; O, secretion by cells perifused with Ca2+-free medium. Each point represents the mean of six (for CRF) or five (for 8-Br-cAMP) determinations; brackets indicate the SEM. *, P < 0.05 [by Student's paired t test, vs. basal secretion (mean IR-ACTH content of the five fractions collected during - 4 through 0 min)].
rate of increase in ACTH secretion during the initial incremental phase of the response to CRF (9, 14) was identical in complete and in Ca2+-free medium, but the initial rate of increase after stimulation with 8-Br-cAMP was slower in Ca2+-free medium (Fig. 3). Unless removing Ca2+ from the medium slows the diffusion of 8-Br-cAMP across the membrane, this difference suggests that CRF may act via a mechanism other than activation of cAMPdependent protein kinase-A to evoke the initial incremental ACTH secretory response. Effect of Ca2+ channel blockade on the response to CRF
Others have reported that Ca2+ channel blockers, including nimodipine, reduce the ACTH secretory response to CRF (5-7). We duplicated these results in monolayer cultures of rat anterior pituitary cells (data not shown), but were surprised to find that addition of up to 0.2 mM nimodipine to complete medium for periods as long as 1 h had no effect on the response to CRF in the microperifusion system. We considered that light might be inactivating the agent, so the entire apparatus, excluding the
Endo • 1990 Voll26«No2
fraction collector, was covered with aluminum foil. There was still no effect (data not shown). Therefore, the following experiment was performed to expose the cells to nimodipine before they were perifused with complete Ca2+-containing medium. Cells in four identical perifusion chambers were exposed simultaneously to four 20min pulses of CRF at intervals of 50 min. The first and last CRF pulses were applied after the cells had been perifused with complete medium for 50 min. The second CRF pulse was applied after the cells had been perifused with Ca2+-free medium for 50 min, and the third pulse was delivered after a 20-min perifusion with Ca2+-free medium containing nimodipine, followed by a 30-min perifusion with complete medium containing the same 0.2-mM concentration of nimodipine. As in previous experiments, removal of Ca2+e reduced the ACTH secretory response to CRF by 74% (303 ± 21 vs. 1168 ± 27 pg/20 min; P < 0.001; Fig. 4). Addition of nimodipine to complete medium inhibited the response by 40% (698 ± 99 vs. 1168 ± 27 pg/20 min; P < 0.001). Furthermore, the response to CRF 50 min after removal of nimodipine from the medium was reduced by 27% (855 ± 90 vs. 1168 ± 27 pg/20 min; P < 0.001), suggesting that nimodipine had a prolonged effect after it had been removed from the perifusion medium. To examine further the effects of nimodipine, six identical perifusion chambers were randomly assigned to one of two groups. The first two CRF pulses were applied to both groups under the same conditions as previously. For the third pulse, however, the chambers in one group were perfused with nimodipine in Ca2+-free medium for 20 min and then with nimodipine in complete medium for 30 min, whereas the other group was perfused with Ca2+-FREE
120 -
action
.2 o 2
140 -i CRF
NIMODIPINE
100 80 -
Q.
I X i—
R-AC
852
60 40 20 0
J
0
10
20
70
80
90
110 140 150 160 210 220 230
TIME - min
FIG. 4. Effect of nimodipine, a Ca2+ channel blocker, on the ACTH secretory response to CRF. Four identical chambers were each loaded with 0.7 x 106 4-day cultured cells and were perfused with 10 nM CRF for 20 min at 50-min intervals on four occasions: in complete medium, in Ca2+-free medium, in Ca2+-free medium containing 0.2 mM nimodipine, and again in complete medium. The boxes indicate the periods during which the cells were exposed to CRF, Ca2+-free medium, and nimodipine. IR-ACTH, Immunoreactive ACTH. Each point represents the mean of four determinations; brackets indicate the SEM.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 04:54 For personal use only. No other uses without permission. . All rights reserved.
Effect of depleting Ca2+i on responses to CRF and 8-BrcAMP Depletion of Ca2+; reduced the integrated ACTH response to CRF by 70% (648 ± 34 vs. 2147 ± 245 pg/20 min; P < 0.01) and the response to 8-Br-cAMP by 71% (504 ± 5 vs. 1757 ± 45 pg/20 min; P < 0.001; Fig. 6). As when Ca2+ was removed from the medium, the rate of increase in ACTH secretion was slower after 8-Br-cAMP in Ca2+-depleted cells. However, in contrast to when Ca2+ 1 ||
140 -i .2 0
Ca»-FREE~Ti NIMODIPINE |
12
