0013-7227/90/1262-0809$02.00/0 Endocrinology Copyright© 1990 by The Endocrine Society
Vol. 126, No. 2 Printed in U.S.A.
Effect of Secretagogues on Components of the Secretory System in AtT-20 Cells* ELIZABETH A. THIELE AND BETTY A. EIPPER Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205
ABSTRACT. The mouse corticotrope tumor cell line AtT-20/ D16v was used to investigate the effects of chronic treatment with various secretagogues on individual components of the secretory pathway. Secretagogues acting in part through receptors linked to guanine nucleotide-binding regulatory proteins [CRF and somatostatin (SS)] and agents by-passing membrane receptors (phorbol myristate acetate and dexamethasone) were examined. Effects on the secretory product were evaluated by measuring levels of pro-ACTH/endorphin mRNA and hormone secretion. Effects on posttranslational processing enzymes were evaluated by measuring levels of the mRNAs encoding carboxypeptidase-E and peptidyl-glycine-a-amidating monooxygenase (PAM); cellular levels of PAM activity were also measured. The mRNAs encoding the G-proteins in AtT-20 cells were identified, and secretagogue effects on the G-protein signal transduction system were evaluated by measuring levels of the mRNAs en-
coding (as, a)i2, and /32. No single parameter adequately characterizes the regulatory state of the complex secretory apparatus. Although levels of pro-ACTH/endorphin (PAE) mRNA accurately reflected hormone secretion after chronic CRF or dexamethasone treatment, chronic SS treatment elevated PAE mRNA levels in the face of reduced hormone secretion. Levels of PAM mRNA generally changed in parallel with levels of PAE mRNA; in contrast, levels of carboxypeptidase-E mRNA were unaffected by any of the secretagogues tested. Secretagogues acting through distinct G-proteins (CRF and SS) as well as dexamethasone brought about a coordinate increase in the level of the mRNAs encoding the three G-protein subunits examined. Treatment with phorbol myristate acetate caused a slight decrease in the levels of the G-protein subunit mRNAs. {Endocrinology 126: 809-817, 1990)
M
sponse to /3-adrenergic agents, while corticotropes exhibit only a transient stimulatory response. The effect of a given secretagogue on the secretory process can be approached by examining specific steps in the secretory process or by screening the effect of that secretagogue on all cellular mRNAs or proteins. We have chosen to use the AtT-20/D16v mouse pituitary corticotrope tumor cell line to examine the effects of secretagogue treatment on a variety of specific steps in the secretory pathway. The AtT-20 cells provide a uniform cell population responsive to agents acting through membrane receptors and guanine nucleotide-binding regulatory proteins to stimulate or inhibit hormone secretion. In AtT-20 cells, CRF stimulates secretion via cell surface receptors coupled through Ga to adenylate cyclase and increases PAE gene transcription, mRNA levels, and biosynthesis (8-10). Somatostatin (SS) inhibits PAE secretion through a dual mechanism; it inhibits secretagogue-stimulated cAMP production by acting through G; and inhibits cAMP- independent hormone release through a reduction in intracellular free Ca2+ (11-14). The effects of SS treatment on PAE mRNA levels have not been previously reported. Hormone secretion can be inhibited by glucocorticoids, which act at multiple sites, including PAE gene transcription, CRF receptor level, and secretion itself (10, 15-18). Hormone secretion can
ANY of the acute effects of secretagogues on peptide-secreting target cells have been elucidated in great detail. The interaction of ligands with cell surface receptors, the role of guanine nucleotide-binding regulatory proteins in signal transduction, the generation of various intracellular signals, and the release of stored products from their vesicular storage sites are all being characterized (1-5). We have been concerned with the chronic effects of secretagogues on target cells (6, 7). The production and secretion of peptide products involve components of the rough endoplasmic reticulum, the Golgi apparatus, and secretory granules. It is not yet clear which components are normally rate limiting and which are subject to regulation by chronic secretagogue treatment. Corticotropes exposed to CRF for 2 weeks continue to respond with increased synthesis and secretion of pro-ACTH/endorphin, while melanotropes, which synthesize the same prohormone, exhibit only a transient stimulatory response when exposed to CRF. In contrast, melanotropes are capable of a sustained reReceived August 7, 1989. Address all correspondence and requests for reprints to: Dr. Betty A. Eipper, Department of Neuroscience, Johns Hopkins University School of Medicine, 725 North Wolfe Street, Baltimore, Maryland 21205. * This work was supported by NIDA Grants DA-00266 and DA00098 (to B.A.E.) and NIH Grant GM-07309 (to E.A.T.).
809
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM
810
also be stimulated by phorbol esters, acting via their effects on protein kinase-C (8, 19-22). Although secretagogue effects on peptide hormone secretion are often closely linked to parallel effects on prohormone synthesis, the two processes can be regulated independently in other cell types (23-25). Therefore, we monitored the effects of CRF, SS, dexamethasone, and phorbol rriyristate acetate (PMA) treatment on pro-ACTH/endorphin mRNA levels and hormone secretion. We then examined the effect of each secretagogue on two secretory granule-associated enzymes known to be involved in peptide hormone production, carboxypeptidase-E (CPE; also carboxypeptidase-H and enkephalin convertase; EC 3.4.17.10) (26) and peptidylglycine a-amidating monooxygenase (PAM; EC 1.14.17.3) (27, 28). To examine the effect of secretagogues on components of the signal transduction system, we determined the effect of each secretagogue on the major guanine nucleotide-binding regulatory protein subunits in AtT-20 cells. Each secretagogue exerts a unique combination of effects on these individual components of the secretory system. Materials and Methods Cell culture and immunoassay Monolayers of AtT-20/D16v cells were grown in Dulbecco's Modified Eagle's Medium-Ham's F-12 medium (DMEM/F12; 1:1) containing 10% horse serum, 10% fetal bovine serum, and 10% Nu Serum (Collaborative Research, Lexington, MA) as described previously (29). Growing cells in charcoal-stripped serum for 2 weeks had no significant effect on their responsiveness to CRF and dexamethasone. Cells were exposed for 6 h to control medium or to medium containing 100 nM CRF [human CRF-(1-41)NH2; Peninsula Laboratories, Inc., Belmont, CA] or 100 nM somatostatin (Sigma Chemical Co., St. Louis, MO) or for 3 h to medium containing 10 nM PMA (Sigma Chemical Co.). Cells were treated with dexamethasone by growth in medium containing 1 nM dexamethasone (Lypho-Med, Inc., Chicago, IL) for 72 h. To mimic the feeding schedule of cells treated acutely with CRF, SS, or PMA, these dexamethasonepretreated cells were fed with fresh medium containing 1 /xM dexamethasone 6 h before harvesting. Mixtures of secretagogues were not evaluated in these studies. Doses of secretagogues were selected to give a maximal effect on secretion (9, 11, 12, 19, 29). At the end of the incubation time, medium was removed, and cells were extracted for immunoassay and protein determination as described previously (29). Secretion of PAE-derived peptides and cellular PAE content were determined by immunoassay with an antiserum to the amino-terminal region of PAE (16K fragment) that recognizes hormone precursor as well as products (29). Secreted peptides were normalized to cell protein content, and secretion rates are expressed as picomoles of 16K fragment immunoactivity secreting per ng cell protein/h. Cells were washed once with serum-free DMEM/F12, gently
Endo • 1990 Vol 126 • No 2
scraped from the plates with a rubber policeman, and pelleted by centrifugation. Cell pellets were homogenized in 20 mM NaTES, pH 7.4, containing 10 mM mannitol, phenylmethylsulfonylfluoride (final concentration, 0.3 mg/ml), leupeptin (2 jug/ml), benzamidine (16 /ig/ml), and soybean trypsin inhibitor (10 jiig/ml). Soluble and membrane fractions were prepared as previously described (30). Washed membranes were resuspended in a minimal volume of 20 mM NaTES, pH 7.4, containing 10 mM mannitol, 1% Triton X-100, and protease inhibitors. Soluble and membrane fractions were assayed for PAM activity using D-Tyr-Val-Gly as substrate (29). PAM activity, in picomoles of product per mg protein/h, represents the mean of assay quadruplicates (two amounts of sample, each assayed in duplicate), which varied by less than 15%. Total specific activity was calculated by summing total activity in soluble and membrane fractions and normalizing to total protein. RNA isolation and characterization Total cellular RNA was isolated from AtT-20 cells by guanidine isothiocyanate extraction and CsCl centrifugation as described previously (29). Yields of total RNA, determined by measuring A26o, were approximately 40-75 ng total RNA/mg cell protein. Duplicate aliquots of total RNA (10 /xg) were fractionated on 1.5% agarose-6% formaldehyde denaturing gels, transferred to Nytran (Schleicher and Schuell, Keene, NH) or Nitrocellulose Plus (Micron Separations, Inc., Westboro, MA) by capillary blotting, and prehybridized as described previously (29). Filters were hybridized for 18 h with 1 x 106 cpm denatured nick-translated cDNA/ml solution; nick-translated probes had specific activities of 0.5-1.0 X 109 cpm/jtg DNA. Filters were washed once for 30 min in 2 x SSC (3. M NaCl, 0.3 M Nacitrate, pH 7.0) 0.1% sodium dodecyl sulfate (SDS) and four times for 30 min each in 0.1 x SSC-0.1% SDS at 50 C, then exposed to x-ray film with an intensifying screen at -70 C. To perform successive hybridizations, probes were removed from filters by incubation in 50% formamide-1 X SSC at 70 C for 60 min (Nytran) or by boiling for 20 min in 0.01 X SSC-1.0% SDS (Nitrocellulose Plus). The washed filters were checked by exposure to x-ray film overnight. Filters were hybridized up to five times with no apparent loss in signal intensity. Preparation of PAE and PAM cDNA probes was described previously (29). The rat CPE cDNA was kindly provided by Dr. L. Fricker (31); a 2-kilobase (kb) EcoRI fragment containing most of the translated portion of CPE and a small amount of the 3'-untranslated region was removed from the pSP-65 vector. The G-protein cDNAs were kindly provided by Drs. D. Jones and R. Reed (The Johns Hopkins University School of Medicine) and correspond to approximately full-length rat olfactory epithelium mRNAs (32). The interspecies nucleotide homology of the G-proteins for rat and mouse is very high (as, 98%; a;, 96%) (32). G-Protein cDNA fragments were removed from the appropriate pGEM-2 plasmid by digestion with EcoRl. RNA quantitation All blots were stripped and hybridized with a frog ribosomal cDNA (29) or cyclophilin (1B15) cDNA (kindly provided by
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM Dr. J. G. Sutcliffe, Research Institute, Scripps Clinic) in order to correct for variations in the amount of total RNA transferred to the filter; cyclophilin mRNA (~1 kb) is constitutively expressed in many tissues and encodes a cyclosporin-A-binding protein (33). Integrated optical densities of individual bands on autoradiograms were converted to apparent disintegrations per min by comparison to a standard curve (29). Standards were prepared by applying serial dilutions of nick-translated cDNA to nitrocellulose using a slot blot apparatus (Schleicher and Schuell); standards were exposed along with Northern blots on each autoradiogram. After treatment with the various secretagogues, the densitized levels of 18S RNA or 1B15 mRNA per ng RNA quantitated by optical density were constant. Therefore, to increase the accuracy of measurements, data for PAE, PAM, CPE, and the G-protein mRNA species were normalized to the amount of 18S RNA or 1B15 mRNA present on the blot; similar results were obtained using 18S RNA or IB 15 RNA for normalization. Duplicate lanes were averaged, and secretagogue effects are all reported relative to values in control cells. Errors were calculated from normalized values (to 18S RNA or IB 15 mRNA) from the different experiments; the average SEM in measurements of RNA ratios was ±9%. Statistical significance was determined by Student's t test.
Establishing secretagogue effects on PAE mRNA levels and PAE secretion The effects of CRF, SS, PMA and dexamethasone on the rate of hormone secretion were compared to their effects on PAE mRNA levels (Figs. 1 and 2). Doses of secretagogue that produced maximal effects on secretion were compared (9, 11,12,19, 29, 34). As expected, treat1
mRNA
.
-
1
250
.
-
1
200
.
100
-
1
150
--•
- Bjj
50 -
CTRL
I
1 i• 1 I *•
I
CRF
Secretion
*
i• *
. * Bif PMA
Dex
ment with CRF or dexamethasone produced parallel changes in levels of PAE mRNA and hormone secretion. In contrast, treatment with PMA or SS failed to produce parallel changes in PAE mRNA levels and secretion (Figs. 1 and 2). The ability of PMA to stimulate secretion undergoes down-regulation in pituitary corticotropes (6, 35), and a similar down-regulation of the secretory response to PMA was seen with AtT-20 cells (Fig. 2A). PMA stimulated PAE secretion from AtT-20 cells greater than 2-fold for up to 4 h; longer incubation times failed to maintain this stimulation. A slight increase in PAE mRNA levels (113 ± 4%) was detected after treatment with PMA for 3 h; even when treatment times were extended to 12 or 24 h, PMA had only a slight effect on PAE mRNA levels (107% and 104% of control, respectively; Fig. 2B). AtT-20 cells were exposed to 100 nM SS, which gives a maximal effect on cAMP and Ca2+ levels in AtT-20 cells and inhibits basal and stimulated PAE secretion (11, 12, 14). Although SS treatment reduced hormone secretion as expected, PAE mRNA levels were elevated to 135 ± 11% of control values (Figs. 1 and 2C). Secretagogue effects on posttranslational processing enzymes
Results
300
811
SS
FIG. 1. Comparison of secretagogue effects on PAE secretion and mRNA levels. The amount of PAE-derived peptide in the medium of secretagogue-treated AtT-20 cells was determined by 16K fragment RIA (picomoles of 16K fragment immunoactivity secreted per ng cell protein/h). The amount of PAE mRNA was determined by quantitative Northern analysis. Secretagogue effects on secretion and mRNA levels are expressed relative to control values (CTRL) and represent results from four independent experiments. Dex, Dexamethasone. Error bars indicate the SEM of the normalized values. *, Significantly different from control, P < 0.05.
CPE and PAM each catalyze late reactions in the conversion of prohormones into product peptides and are the only mammalian prohormone posttranslational processing enzymes that have been purified and cloned. Both enzymes are localized primarily to secretory granules and occur in membrane-associated and soluble forms (26, 28). The effect of each of the secretagogues on levels of CPE and PAM mRNA was determined (Figs. 3 and 4). As shown previously, CRF and dexamethasone both regulated levels of PAM mRNA in parallel with levels of PAE mRNA and secretion (27, 29). In contrast, levels of CPE mRNA were unaltered by CRF or dexamethasone treatment (Fig. 3). Just as PMA treatment resulted in only a slight change in the levels of PAE mRNA, treatment with PMA produced only a slight change in the levels of PAM mRNA and no change in the levels of CPE mRNA. Continuation of PMA treatment for 24 h failed to change PAM or CPE mRNA levels (data not shown). The effect of SS on levels of PAM mRNA paralleled its effect on levels of PAE mRNA (Figs. 3 and 4). Although SS decreased hormone secretion, it brought about a coordinate increase in levels of PAM and PAE mRNA. As with the other secretagogues tested, levels of CPE mRNA were unaffected by SS treatment (100 ± 12%). Levels of mRNAs encoding individual secretory granule constituents are clearly subject to separate regulatory processes. Enzyme assays capable of measuring the level of activ-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM
812
0.16 —
0.12-
0.08-
Endo • 1990 Vol 126 • No 2
ft
• 0.04-
n
. , , I . , . I , , , I
12
, , I , , , I
16
20
. , I . , , I , . . I • , , I .
24
28
32
36
. I . , , I , , , I ,
40
44
48
Treatment (hours)
3h
12h
control control
PAE
SS
•••••i PAE
18S
B 1B15
FlG. 2. Effects of PMA and SS on PAE secretion and mRNA levels. A, Duplicate wells of AtT-20 cells were treated with 10 nM PMA for the times indicated. Media in all the wells were replaced 48 h before extraction to avoid feeding artifacts, and PMA was added to each well from a concentrated stock at the appropriate time. Secretion of PAE-derived peptides over the final 3-h period was measured by replacing medium with fresh medium containing the same concentration of PMA for the final 3-h period; for the 1 h time point, media were changed 3 h before extraction, and PMA was added 1 h before extraction. Therefore, for all time points, the cells were extracted at the same time. Data are presented as picomoles of immunoreactive 16K fragment secreted per fig cell protein/h. The data represent the average of the duplicate wells from a single experiment, which differed by less than 10%; similar results were obtained in three separate experiments. B, AtT-20 cells were treated with 10 nM PMA for 3 and 12 h. Levels of PAE mRNA (~1 kb) were determined by Northern analysis. The blot was hybridized with the PAE cDNA probe (top) and then stripped and hybridized with the ribosomal cDNA probe (bottom). C, Total cellular RNA isolated from control and SS-treated AtT-20 cells was subjected to Northern analysis. The blot was hybridized with PAE cDNA (top) and then stripped and hybridized with the 1B15 cDNA probe (bottom). The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM
160
-
140
-
120
-
813
activity of soluble and membrane-associated PAM changed in a similar fashion in response to all secretagogues, with a slightly larger effect on soluble activity than on particulate activity. As with tissue hormone content, measurement of the tissue content of processing enzyme activity alone is a poor indicator of the effects of regulatory factors impinging on that tissue; PAM specific activity was decreased in response to agents that both increased and decreased levels of PAM mRNA.
PAE PAM CPE
100 —
Secretagogue effects on components of the signal transdution system: G-proteins Ctrl
CRF
PMA
Dex
SS
FIG. 3. Comparison of secretagogue effects on PAE, PAM and CPE mRNA levels. Secretagogue effects on PAM and CPE mRNA levels relative to control values (Ctrl) were obtained from Northern analyses as described in Fig. 4. Data are from three separate experiments. Data for PAE mRNA levels are shown again for comparison. Dex, Dexamethasone. The SEM was calculated from normalized values and is represented by error bars. * , Significantly different from control, P < 0.05.
ity of these secretory granule-associated enzymes in crude subcellular fractions have been developed (30). Therefore, the effect of secretagogue treatment on levels of enzyme activity could be determined. As demonstrated previously, treatment of AtT-20 cells with either CRF or dexamethasone resulted in a decrease in the total specific activity of PAM (29, 34, 36). CPE activity was previously shown to be unresponsive to CRF or dexamethasone treatment (34, 36). Despite the fact that PMA stimulated hormone secretion and decreased the cellular content of PAE, treatment with PMA failed to alter the total specific activity of PAM in AtT-20 cells (Fig. 5). SS treatment, while increasing levels of PAM mRNA and inhibiting hormone secretion, failed to increase the specific activity of PAM in AtT-20 cells (Fig. 5). The specific PMA FIG. 4. Analysis of CPE and PAM mRNA levels after secretagogue treatment. A, Total cellular RNA from PMAand SS-treated and control (ctrl) AtT20 cells (10 fig) was subjected to Northern analysis; the blots were hybridized with nick-translated PAM cDNA and then stripped and rehybridized with nick-translated 1B15 cDNA. B, Total cellular RNA (10 ng) from control AtT20 cells or cells treated with SS was subjected to Northern analysis. Blots were hybridized with nick-translated CPE cDNA, stripped, and rehybridized with nick-translated 1B15 cDNA.
Although many secretagogues are known to act on AtT-20 cells via G-proteins, the specific G-proteins present in AtT-20 cells have not been characterized. Total cellular RNA isolated from control AtT-20 cells was subjected to Northern analysis, and filters were hybridized with cDNA probes encoding all of the identified species of as, a{, and /3 (Fig. 6) (2, 32, 37, 38). The as and ai2 cDNA probes each hybridized to a single major mRNA species, with apparent sizes of 1.7 and 2.2 kb, respectively. Hybridization with the /?i cDNA probe identified two mRNA species of 3.0 and 1.6 kb; the intensity of hybridization of the 3.0-kb form was generally 2-fold less than that of the 1.6-kb form. The aa cDNA probe hybridized to a single mRNA species with an apparent size of 3.3 kb. The ai3 cDNA probe recognized mRNA species with apparent sizes of 3.2 and 2.2 kb; the 2.2 kb mRNA represents minor cross-hybridization of the
Vol. 126, No. 2 Printed in U.S.A.
Effect of Secretagogues on Components of the Secretory System in AtT-20 Cells* ELIZABETH A. THIELE AND BETTY A. EIPPER Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205
ABSTRACT. The mouse corticotrope tumor cell line AtT-20/ D16v was used to investigate the effects of chronic treatment with various secretagogues on individual components of the secretory pathway. Secretagogues acting in part through receptors linked to guanine nucleotide-binding regulatory proteins [CRF and somatostatin (SS)] and agents by-passing membrane receptors (phorbol myristate acetate and dexamethasone) were examined. Effects on the secretory product were evaluated by measuring levels of pro-ACTH/endorphin mRNA and hormone secretion. Effects on posttranslational processing enzymes were evaluated by measuring levels of the mRNAs encoding carboxypeptidase-E and peptidyl-glycine-a-amidating monooxygenase (PAM); cellular levels of PAM activity were also measured. The mRNAs encoding the G-proteins in AtT-20 cells were identified, and secretagogue effects on the G-protein signal transduction system were evaluated by measuring levels of the mRNAs en-
coding (as, a)i2, and /32. No single parameter adequately characterizes the regulatory state of the complex secretory apparatus. Although levels of pro-ACTH/endorphin (PAE) mRNA accurately reflected hormone secretion after chronic CRF or dexamethasone treatment, chronic SS treatment elevated PAE mRNA levels in the face of reduced hormone secretion. Levels of PAM mRNA generally changed in parallel with levels of PAE mRNA; in contrast, levels of carboxypeptidase-E mRNA were unaffected by any of the secretagogues tested. Secretagogues acting through distinct G-proteins (CRF and SS) as well as dexamethasone brought about a coordinate increase in the level of the mRNAs encoding the three G-protein subunits examined. Treatment with phorbol myristate acetate caused a slight decrease in the levels of the G-protein subunit mRNAs. {Endocrinology 126: 809-817, 1990)
M
sponse to /3-adrenergic agents, while corticotropes exhibit only a transient stimulatory response. The effect of a given secretagogue on the secretory process can be approached by examining specific steps in the secretory process or by screening the effect of that secretagogue on all cellular mRNAs or proteins. We have chosen to use the AtT-20/D16v mouse pituitary corticotrope tumor cell line to examine the effects of secretagogue treatment on a variety of specific steps in the secretory pathway. The AtT-20 cells provide a uniform cell population responsive to agents acting through membrane receptors and guanine nucleotide-binding regulatory proteins to stimulate or inhibit hormone secretion. In AtT-20 cells, CRF stimulates secretion via cell surface receptors coupled through Ga to adenylate cyclase and increases PAE gene transcription, mRNA levels, and biosynthesis (8-10). Somatostatin (SS) inhibits PAE secretion through a dual mechanism; it inhibits secretagogue-stimulated cAMP production by acting through G; and inhibits cAMP- independent hormone release through a reduction in intracellular free Ca2+ (11-14). The effects of SS treatment on PAE mRNA levels have not been previously reported. Hormone secretion can be inhibited by glucocorticoids, which act at multiple sites, including PAE gene transcription, CRF receptor level, and secretion itself (10, 15-18). Hormone secretion can
ANY of the acute effects of secretagogues on peptide-secreting target cells have been elucidated in great detail. The interaction of ligands with cell surface receptors, the role of guanine nucleotide-binding regulatory proteins in signal transduction, the generation of various intracellular signals, and the release of stored products from their vesicular storage sites are all being characterized (1-5). We have been concerned with the chronic effects of secretagogues on target cells (6, 7). The production and secretion of peptide products involve components of the rough endoplasmic reticulum, the Golgi apparatus, and secretory granules. It is not yet clear which components are normally rate limiting and which are subject to regulation by chronic secretagogue treatment. Corticotropes exposed to CRF for 2 weeks continue to respond with increased synthesis and secretion of pro-ACTH/endorphin, while melanotropes, which synthesize the same prohormone, exhibit only a transient stimulatory response when exposed to CRF. In contrast, melanotropes are capable of a sustained reReceived August 7, 1989. Address all correspondence and requests for reprints to: Dr. Betty A. Eipper, Department of Neuroscience, Johns Hopkins University School of Medicine, 725 North Wolfe Street, Baltimore, Maryland 21205. * This work was supported by NIDA Grants DA-00266 and DA00098 (to B.A.E.) and NIH Grant GM-07309 (to E.A.T.).
809
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM
810
also be stimulated by phorbol esters, acting via their effects on protein kinase-C (8, 19-22). Although secretagogue effects on peptide hormone secretion are often closely linked to parallel effects on prohormone synthesis, the two processes can be regulated independently in other cell types (23-25). Therefore, we monitored the effects of CRF, SS, dexamethasone, and phorbol rriyristate acetate (PMA) treatment on pro-ACTH/endorphin mRNA levels and hormone secretion. We then examined the effect of each secretagogue on two secretory granule-associated enzymes known to be involved in peptide hormone production, carboxypeptidase-E (CPE; also carboxypeptidase-H and enkephalin convertase; EC 3.4.17.10) (26) and peptidylglycine a-amidating monooxygenase (PAM; EC 1.14.17.3) (27, 28). To examine the effect of secretagogues on components of the signal transduction system, we determined the effect of each secretagogue on the major guanine nucleotide-binding regulatory protein subunits in AtT-20 cells. Each secretagogue exerts a unique combination of effects on these individual components of the secretory system. Materials and Methods Cell culture and immunoassay Monolayers of AtT-20/D16v cells were grown in Dulbecco's Modified Eagle's Medium-Ham's F-12 medium (DMEM/F12; 1:1) containing 10% horse serum, 10% fetal bovine serum, and 10% Nu Serum (Collaborative Research, Lexington, MA) as described previously (29). Growing cells in charcoal-stripped serum for 2 weeks had no significant effect on their responsiveness to CRF and dexamethasone. Cells were exposed for 6 h to control medium or to medium containing 100 nM CRF [human CRF-(1-41)NH2; Peninsula Laboratories, Inc., Belmont, CA] or 100 nM somatostatin (Sigma Chemical Co., St. Louis, MO) or for 3 h to medium containing 10 nM PMA (Sigma Chemical Co.). Cells were treated with dexamethasone by growth in medium containing 1 nM dexamethasone (Lypho-Med, Inc., Chicago, IL) for 72 h. To mimic the feeding schedule of cells treated acutely with CRF, SS, or PMA, these dexamethasonepretreated cells were fed with fresh medium containing 1 /xM dexamethasone 6 h before harvesting. Mixtures of secretagogues were not evaluated in these studies. Doses of secretagogues were selected to give a maximal effect on secretion (9, 11, 12, 19, 29). At the end of the incubation time, medium was removed, and cells were extracted for immunoassay and protein determination as described previously (29). Secretion of PAE-derived peptides and cellular PAE content were determined by immunoassay with an antiserum to the amino-terminal region of PAE (16K fragment) that recognizes hormone precursor as well as products (29). Secreted peptides were normalized to cell protein content, and secretion rates are expressed as picomoles of 16K fragment immunoactivity secreting per ng cell protein/h. Cells were washed once with serum-free DMEM/F12, gently
Endo • 1990 Vol 126 • No 2
scraped from the plates with a rubber policeman, and pelleted by centrifugation. Cell pellets were homogenized in 20 mM NaTES, pH 7.4, containing 10 mM mannitol, phenylmethylsulfonylfluoride (final concentration, 0.3 mg/ml), leupeptin (2 jug/ml), benzamidine (16 /ig/ml), and soybean trypsin inhibitor (10 jiig/ml). Soluble and membrane fractions were prepared as previously described (30). Washed membranes were resuspended in a minimal volume of 20 mM NaTES, pH 7.4, containing 10 mM mannitol, 1% Triton X-100, and protease inhibitors. Soluble and membrane fractions were assayed for PAM activity using D-Tyr-Val-Gly as substrate (29). PAM activity, in picomoles of product per mg protein/h, represents the mean of assay quadruplicates (two amounts of sample, each assayed in duplicate), which varied by less than 15%. Total specific activity was calculated by summing total activity in soluble and membrane fractions and normalizing to total protein. RNA isolation and characterization Total cellular RNA was isolated from AtT-20 cells by guanidine isothiocyanate extraction and CsCl centrifugation as described previously (29). Yields of total RNA, determined by measuring A26o, were approximately 40-75 ng total RNA/mg cell protein. Duplicate aliquots of total RNA (10 /xg) were fractionated on 1.5% agarose-6% formaldehyde denaturing gels, transferred to Nytran (Schleicher and Schuell, Keene, NH) or Nitrocellulose Plus (Micron Separations, Inc., Westboro, MA) by capillary blotting, and prehybridized as described previously (29). Filters were hybridized for 18 h with 1 x 106 cpm denatured nick-translated cDNA/ml solution; nick-translated probes had specific activities of 0.5-1.0 X 109 cpm/jtg DNA. Filters were washed once for 30 min in 2 x SSC (3. M NaCl, 0.3 M Nacitrate, pH 7.0) 0.1% sodium dodecyl sulfate (SDS) and four times for 30 min each in 0.1 x SSC-0.1% SDS at 50 C, then exposed to x-ray film with an intensifying screen at -70 C. To perform successive hybridizations, probes were removed from filters by incubation in 50% formamide-1 X SSC at 70 C for 60 min (Nytran) or by boiling for 20 min in 0.01 X SSC-1.0% SDS (Nitrocellulose Plus). The washed filters were checked by exposure to x-ray film overnight. Filters were hybridized up to five times with no apparent loss in signal intensity. Preparation of PAE and PAM cDNA probes was described previously (29). The rat CPE cDNA was kindly provided by Dr. L. Fricker (31); a 2-kilobase (kb) EcoRI fragment containing most of the translated portion of CPE and a small amount of the 3'-untranslated region was removed from the pSP-65 vector. The G-protein cDNAs were kindly provided by Drs. D. Jones and R. Reed (The Johns Hopkins University School of Medicine) and correspond to approximately full-length rat olfactory epithelium mRNAs (32). The interspecies nucleotide homology of the G-proteins for rat and mouse is very high (as, 98%; a;, 96%) (32). G-Protein cDNA fragments were removed from the appropriate pGEM-2 plasmid by digestion with EcoRl. RNA quantitation All blots were stripped and hybridized with a frog ribosomal cDNA (29) or cyclophilin (1B15) cDNA (kindly provided by
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM Dr. J. G. Sutcliffe, Research Institute, Scripps Clinic) in order to correct for variations in the amount of total RNA transferred to the filter; cyclophilin mRNA (~1 kb) is constitutively expressed in many tissues and encodes a cyclosporin-A-binding protein (33). Integrated optical densities of individual bands on autoradiograms were converted to apparent disintegrations per min by comparison to a standard curve (29). Standards were prepared by applying serial dilutions of nick-translated cDNA to nitrocellulose using a slot blot apparatus (Schleicher and Schuell); standards were exposed along with Northern blots on each autoradiogram. After treatment with the various secretagogues, the densitized levels of 18S RNA or 1B15 mRNA per ng RNA quantitated by optical density were constant. Therefore, to increase the accuracy of measurements, data for PAE, PAM, CPE, and the G-protein mRNA species were normalized to the amount of 18S RNA or 1B15 mRNA present on the blot; similar results were obtained using 18S RNA or IB 15 RNA for normalization. Duplicate lanes were averaged, and secretagogue effects are all reported relative to values in control cells. Errors were calculated from normalized values (to 18S RNA or IB 15 mRNA) from the different experiments; the average SEM in measurements of RNA ratios was ±9%. Statistical significance was determined by Student's t test.
Establishing secretagogue effects on PAE mRNA levels and PAE secretion The effects of CRF, SS, PMA and dexamethasone on the rate of hormone secretion were compared to their effects on PAE mRNA levels (Figs. 1 and 2). Doses of secretagogue that produced maximal effects on secretion were compared (9, 11,12,19, 29, 34). As expected, treat1
mRNA
.
-
1
250
.
-
1
200
.
100
-
1
150
--•
- Bjj
50 -
CTRL
I
1 i• 1 I *•
I
CRF
Secretion
*
i• *
. * Bif PMA
Dex
ment with CRF or dexamethasone produced parallel changes in levels of PAE mRNA and hormone secretion. In contrast, treatment with PMA or SS failed to produce parallel changes in PAE mRNA levels and secretion (Figs. 1 and 2). The ability of PMA to stimulate secretion undergoes down-regulation in pituitary corticotropes (6, 35), and a similar down-regulation of the secretory response to PMA was seen with AtT-20 cells (Fig. 2A). PMA stimulated PAE secretion from AtT-20 cells greater than 2-fold for up to 4 h; longer incubation times failed to maintain this stimulation. A slight increase in PAE mRNA levels (113 ± 4%) was detected after treatment with PMA for 3 h; even when treatment times were extended to 12 or 24 h, PMA had only a slight effect on PAE mRNA levels (107% and 104% of control, respectively; Fig. 2B). AtT-20 cells were exposed to 100 nM SS, which gives a maximal effect on cAMP and Ca2+ levels in AtT-20 cells and inhibits basal and stimulated PAE secretion (11, 12, 14). Although SS treatment reduced hormone secretion as expected, PAE mRNA levels were elevated to 135 ± 11% of control values (Figs. 1 and 2C). Secretagogue effects on posttranslational processing enzymes
Results
300
811
SS
FIG. 1. Comparison of secretagogue effects on PAE secretion and mRNA levels. The amount of PAE-derived peptide in the medium of secretagogue-treated AtT-20 cells was determined by 16K fragment RIA (picomoles of 16K fragment immunoactivity secreted per ng cell protein/h). The amount of PAE mRNA was determined by quantitative Northern analysis. Secretagogue effects on secretion and mRNA levels are expressed relative to control values (CTRL) and represent results from four independent experiments. Dex, Dexamethasone. Error bars indicate the SEM of the normalized values. *, Significantly different from control, P < 0.05.
CPE and PAM each catalyze late reactions in the conversion of prohormones into product peptides and are the only mammalian prohormone posttranslational processing enzymes that have been purified and cloned. Both enzymes are localized primarily to secretory granules and occur in membrane-associated and soluble forms (26, 28). The effect of each of the secretagogues on levels of CPE and PAM mRNA was determined (Figs. 3 and 4). As shown previously, CRF and dexamethasone both regulated levels of PAM mRNA in parallel with levels of PAE mRNA and secretion (27, 29). In contrast, levels of CPE mRNA were unaltered by CRF or dexamethasone treatment (Fig. 3). Just as PMA treatment resulted in only a slight change in the levels of PAE mRNA, treatment with PMA produced only a slight change in the levels of PAM mRNA and no change in the levels of CPE mRNA. Continuation of PMA treatment for 24 h failed to change PAM or CPE mRNA levels (data not shown). The effect of SS on levels of PAM mRNA paralleled its effect on levels of PAE mRNA (Figs. 3 and 4). Although SS decreased hormone secretion, it brought about a coordinate increase in levels of PAM and PAE mRNA. As with the other secretagogues tested, levels of CPE mRNA were unaffected by SS treatment (100 ± 12%). Levels of mRNAs encoding individual secretory granule constituents are clearly subject to separate regulatory processes. Enzyme assays capable of measuring the level of activ-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM
812
0.16 —
0.12-
0.08-
Endo • 1990 Vol 126 • No 2
ft
• 0.04-
n
. , , I . , . I , , , I
12
, , I , , , I
16
20
. , I . , , I , . . I • , , I .
24
28
32
36
. I . , , I , , , I ,
40
44
48
Treatment (hours)
3h
12h
control control
PAE
SS
•••••i PAE
18S
B 1B15
FlG. 2. Effects of PMA and SS on PAE secretion and mRNA levels. A, Duplicate wells of AtT-20 cells were treated with 10 nM PMA for the times indicated. Media in all the wells were replaced 48 h before extraction to avoid feeding artifacts, and PMA was added to each well from a concentrated stock at the appropriate time. Secretion of PAE-derived peptides over the final 3-h period was measured by replacing medium with fresh medium containing the same concentration of PMA for the final 3-h period; for the 1 h time point, media were changed 3 h before extraction, and PMA was added 1 h before extraction. Therefore, for all time points, the cells were extracted at the same time. Data are presented as picomoles of immunoreactive 16K fragment secreted per fig cell protein/h. The data represent the average of the duplicate wells from a single experiment, which differed by less than 10%; similar results were obtained in three separate experiments. B, AtT-20 cells were treated with 10 nM PMA for 3 and 12 h. Levels of PAE mRNA (~1 kb) were determined by Northern analysis. The blot was hybridized with the PAE cDNA probe (top) and then stripped and hybridized with the ribosomal cDNA probe (bottom). C, Total cellular RNA isolated from control and SS-treated AtT-20 cells was subjected to Northern analysis. The blot was hybridized with PAE cDNA (top) and then stripped and hybridized with the 1B15 cDNA probe (bottom). The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 13 September 2015. at 21:33 For personal use only. No other uses without permission. . All rights reserved.
REGULATION OF AtT-20 SECRETORY SYSTEM
160
-
140
-
120
-
813
activity of soluble and membrane-associated PAM changed in a similar fashion in response to all secretagogues, with a slightly larger effect on soluble activity than on particulate activity. As with tissue hormone content, measurement of the tissue content of processing enzyme activity alone is a poor indicator of the effects of regulatory factors impinging on that tissue; PAM specific activity was decreased in response to agents that both increased and decreased levels of PAM mRNA.
PAE PAM CPE
100 —
Secretagogue effects on components of the signal transdution system: G-proteins Ctrl
CRF
PMA
Dex
SS
FIG. 3. Comparison of secretagogue effects on PAE, PAM and CPE mRNA levels. Secretagogue effects on PAM and CPE mRNA levels relative to control values (Ctrl) were obtained from Northern analyses as described in Fig. 4. Data are from three separate experiments. Data for PAE mRNA levels are shown again for comparison. Dex, Dexamethasone. The SEM was calculated from normalized values and is represented by error bars. * , Significantly different from control, P < 0.05.
ity of these secretory granule-associated enzymes in crude subcellular fractions have been developed (30). Therefore, the effect of secretagogue treatment on levels of enzyme activity could be determined. As demonstrated previously, treatment of AtT-20 cells with either CRF or dexamethasone resulted in a decrease in the total specific activity of PAM (29, 34, 36). CPE activity was previously shown to be unresponsive to CRF or dexamethasone treatment (34, 36). Despite the fact that PMA stimulated hormone secretion and decreased the cellular content of PAE, treatment with PMA failed to alter the total specific activity of PAM in AtT-20 cells (Fig. 5). SS treatment, while increasing levels of PAM mRNA and inhibiting hormone secretion, failed to increase the specific activity of PAM in AtT-20 cells (Fig. 5). The specific PMA FIG. 4. Analysis of CPE and PAM mRNA levels after secretagogue treatment. A, Total cellular RNA from PMAand SS-treated and control (ctrl) AtT20 cells (10 fig) was subjected to Northern analysis; the blots were hybridized with nick-translated PAM cDNA and then stripped and rehybridized with nick-translated 1B15 cDNA. B, Total cellular RNA (10 ng) from control AtT20 cells or cells treated with SS was subjected to Northern analysis. Blots were hybridized with nick-translated CPE cDNA, stripped, and rehybridized with nick-translated 1B15 cDNA.
Although many secretagogues are known to act on AtT-20 cells via G-proteins, the specific G-proteins present in AtT-20 cells have not been characterized. Total cellular RNA isolated from control AtT-20 cells was subjected to Northern analysis, and filters were hybridized with cDNA probes encoding all of the identified species of as, a{, and /3 (Fig. 6) (2, 32, 37, 38). The as and ai2 cDNA probes each hybridized to a single major mRNA species, with apparent sizes of 1.7 and 2.2 kb, respectively. Hybridization with the /?i cDNA probe identified two mRNA species of 3.0 and 1.6 kb; the intensity of hybridization of the 3.0-kb form was generally 2-fold less than that of the 1.6-kb form. The aa cDNA probe hybridized to a single mRNA species with an apparent size of 3.3 kb. The ai3 cDNA probe recognized mRNA species with apparent sizes of 3.2 and 2.2 kb; the 2.2 kb mRNA represents minor cross-hybridization of the
