Stimulus-Secretion Coupling in the Developing Exocrine Pancreas: Secretory Responsiveness to Cholecystokinin

We have studied the onset of secretory responsiveness to cholecystokinin (CCK) during development of the rat exocrine pancreas. Although acinar cells of the fetal pancreas (1 d before birth) are filled with zymogen granules containing the secretory protein, a-amylase, the rate of amylase secretion from pancreatic lobutes incubated in vitro was not increased in response to CCK. In contrast, the rate of CCKstimulated amylase discharge from the neonatal pancreas (1 d after birth) was increased fourto eightfold above that of the fetal gland. The postnatal amplification of secretory responsiveness was not associated with an increase in the number or cell surface expression of ~zsI-CCK binding sites. When 125I-CCK-33 binding proteins were analyzed by affinity crosslinking, two proteins of Mr 210,000 and 100,000160,000 were labeled specifically in both fetal and neonatal pancreas. To determine if cell surface receptors for CCK in the fetal pancreas are functional and able to generate a rise in the cytosolic [Ca++], we measured 4sCa÷* et~ux from tracer-loaded lobules. 45Ca++ efllux from both fetal and neonatal pancreas was comparably increased by CCK, indicating CCKinduced Ca ÷+ mobilization and elevated cytosolic [Ca÷÷]. The Ca ++ ionophore A23187 also stimulated the rate of 45Ca++ extrusion from pancreas of both ages. Increased amylase secretion occurred concurrently with A23187-stimulated 45Ca++ efflux in neonatal pancreas, but not in the fetal gland. A23187 in combination with dibutyryl cAMP potentiated amylase release from the neonatal gland, but not from fetal pancreas. Similarly, the protein kinase C activator, phorbol dibutyrate, did not increase the rate of secretion from the fetal gland when added alone or in combination with A23187 or CCK. We suggest that CCK-receptor interaction in the fetal pancreas triggers intracellular Ca ++ mobilization. However, one or more signal transduction events distal to Ca +÷ mobilization have not yet matured. The onset of secretory response to CCK that occurs postnatally may depend on amplification of these transduction events. T HE polypeptide cholecystokinin (CCK) I serves as a major physiologic secretagogue for the exocrine pancreas. This hormone has been shown to interact specifically with a sialoglycoprotein of Mr 85,000 that is localized on the basolateral plasma membrane of the acinar cell of the adult rat pancreas (41, 42). Consequent to CCK binding to its receptor, a cascade of molecular events occurs that culminates in the exocytotic release of secretory granule content. These signal transduction events are not well understood, but include increased degradation of phosphatidylinositol, generation of diacylglycerol (21, 36, 52), and a transient rise in the cytosolic Ca ÷+ concentration ([Ca÷÷]c) (35, 39, 49). The Ca ++ mobilized during hormone stimulation appears to initially come from an intracellular store, i.e., the trigger pool, since the early phase of secretion can 1. Abbreviations used in this paper: [Ca++]c, cytosolic free calcium concentration; CCK, cholecystokinin; CCK-8, CCK COOH-terminal octapeptide; CCK-33, CCK triacontatriapeptide; C-kinase, protein kinase C; dbcAMP, dibutyryl cyclic AMP; KRH, Krebs Ringer's Hepes buffer; MBS, m-maleimidobenzoyl N-hydroxysuccinimide ester; MEM, Eagle's minimum essential medium; STI, soybean trypsin inhibitor. occur in the absence of extracellular Ca ++ (14, 48, 49). Net efflux of Ca ++ from internal pools occurs during the initial phase of stimulated protein discharge; sustained secretion, however, is dependent on extracellular Ca ++, and net influx of Ca ++ occurs during this period (14, 48). Recent evidence suggests that the sustained phase of stimulated secretion is mediated by the activity of the Ca÷+/phospholipid-dependent protein kinase C (C-kinase) (33, 35, 40). We have studied the mechanism of stimulus-secretion coupling by analyzing the onset of CCK responsiveness during rat pancreatic development. Morphogenesis and cytodifferentiation of the fetal pancreas are well-characterized processes. By day 19 of gestation in the rat (i.e., 3 d before birth), the acinar cells contain the mature complement of intracellular organelles (38). By day 21 of gestation (1 d before birth), a complete set of secretory proteins destined for export are synthesized, transported, and packaged into zymogen granules in a manner similar to that of the adult gland (Arvan, P., and A. Chang, manuscript in preparation; references 43 and 54). At this time, the acinar cell cytoplasm is filled with accumulated © The Rockefeller University Press, 0021-9525/86/12/2353/13 $1.00 The Journal of Cell Biology, Volume 103 (No. 6, Pt. 1), Dec. 1986 2353-2365 2353 on July 1, 2017 jcb.rress.org D ow nladed fom zymogen granules (38), and secretory proteins are released basally in vitro (10, 55). However, the expression of a mature cell surface phenotype (as measured by lectin binding) does not begin to occur until day 21 of gestation (30). Furthermore, treatment of pancreatic organ cultures with the thymidine analog, 5-bromodeoxyuridine, dissociates cell surface glycoconjugate differentiation from cytodifferentiation (31). Such observations suggest that the maturation of plasma membrane proteins (perhaps including those involved in CCK action) may not be tightly coupled with the development of the secretory apparatus. Indeed, it has been observed previously that protein discharge in response to caerulein, a CCK analog, is low in the fetal pancreas (10); significant secretory response to stimulation occurs only after birth (28, 55). In the present study we show that the cell surface expression of CCK binding proteins in the fetal pancreas is temporally distinct from the postnatal development of secretory responsiveness to CCK. One possible explanation for the increased secretory responsiveness of the neonatal pancreas to CCK compared with that of the fetal gland is a more efficient or effective coupling of hormone binding to the generation of the intracellular second messengers, [Ca++]c and diacylglycerol. A similar mechanism for increased hormone sensitivity during development has been previously described in several systems in which the maturation of cellular responsiveness involves a more efficient or effective coupling of receptor occupancy to adenylate cyclase (and generation of the second messenger, cAMP) (24, 27). An alternative explanation for the amplified secretory response in the neonatal pancreas is the maturation of transduction events distal to the mobilization of Ca ++ and/or C-kinase activation. To distinguish between these possibilities, we have examined Ca ++ mobilization in response to CCK during pancreatic development. In addition, we investigated the sensitivity of the secretory machinery in developing pancreas to an elevation in [Ca÷+]c produced by addition of a divalent cation ionophore, and/or to activation of C-kinase by phorbol dibutyrate. Our data show that CCK binding to the fetal pancreas 1 d before birth results in Ca ++ mobilization; however, the gland appears unresponsive to the second messenger generated by either the hormone, or the Ca ++ ionophore A23187. Furthermore, combinations of A23187 with phorbol dibutyrate, or the cAMP derivative, dibutyryl cAMP (dbcAMP) did not stimulate the rate of amylase secretion in the fetal gland, in contrast to the effects observed in neonatal and adult pancreas. Preliminary accounts of this work have been presented (7, 8). Materials and Methods Preparation of Tissue Sprague-Dawley rats were obtained from Carom Research Lab Animals (Wayne, NJ) and allowed to feed or nurse freely. Pancreata were dissected from adult rats, defined as male animals weighing 125-150 g (,'-,2-too-old). Neonatal pancreata were dissected and pooled from one or more litters of rats ",,1-d-old. Fetal rats at day 21 of gestation were removed from one or more mothers with timed pregnancies, and their pancreata dissected and pooled. The dissected glands were placed in cold, oxygenated Eagle's minimum essential medium (MEM) with Hank's salts, buffered with Hepes (25 mM), pH 7.4, and trimmed of connective tissue and/or mesenchyme under a binocular microscope. Amylase Discharge Assay Pancreatic lobules were isolated by dissection with fine scissors. In order to wash away debris spilled into the medium from cells damaged during lobule preparation, the tissue was preincubated in '~10 ml oxygenated MEM containing 0.01% (wt/vol) soybean trypsin inhibitor (STI), 0.1% wt/vol BSA for "~30 min at 37°C, and several changes of the medium were made. To initiate the secretion assay, 2 ml oxygenated MEM was added to six lobules (pooled from several fetal or neonatal glands, or dissected from random regions of the adult pancreas) in the presence or absence of a range of CCK COOH-terminal octapeptide (CCK-8) concentrations. All tubes were incubated for 2 h in a 37°C water bath shaking at 120 oscillations/rain. During the course of the experiment, the samples were gassed with 100% 02 at 30-min intervals. Aliquots of medium (1{30 ~1) were removed every 30 min and replaced with an equal volume of fresh medium containing the appropriate dose of hormone. At the end of 2 h, lobules were rinsed rapidly with MEM before being either sonicated or homogenized in 2 ml of 0.02 % vol/vol Triton X-100, 20 mM NaCI, 10 mM Na phosphate, pH 6.9. Samples were frozen at -20°C before assaying for amylase activity according to the method of Bernfeld (2). Using linear regression analysis, we calculated the rates of amylase discharge (percent of total amylase released per minute) for each experiment. Radioiodination of CCK CCK-8, the most potent form of the hormone in stimulating pancreatic secretion (23), was used in the amylase discharge and 45Ca++ efflux experiments, and 1251-CCK-8 was used in the binding experiments. Since at least one free amino group on radiolabeled CCK triacontatriapeptide (CCK-33) is available for reaction with NH2-reactive cross-linking reagents, J2~ICCK-33 was used for the affinity labeling and autoradiography studies. CCK-33 was acylated with t~I-labeled Bolton-Hunter reagent, as described previously (41). x25I-CCK-33 prepared by this method has a specific activity of ,'~500 Ci/mmol, and has been shown to retain full biologic activity (44). Preparation of Pancreatic Membranes Total membranes were prepared by homogenizing the pancreas using 10 updown strokes with a Brendler Teflon pestle homogenizer (0.08-0.15 mm clearance) driven at 2,200 rprn in 10 times volume per tissue wet weight of Krebs-Ringer Hepes (KRH) buffer diluted fourfold with distilled H20. (KRH: 103 mM NaCI, 4.78 mM KCI, 1.16 mM KH2POa, 1 mM MgSO4, 1 mM CaCI2, 0.2% wt/vol BSA, 25 mM Hepes, pH 7.4.) Diluted KRH buffer (final Mg ++ concentration, 0.25 mM) was used as the homogenization medium to minimize the aggregation of membranes that occurs in the presence of a higher Mg ÷+ concentration (32). Included in the buffer were the protease inhibitors, aprotinin (10 U/ml), leupeptin (10 ~tg/ml), pepstatin (4 p,g/ml), bacitracin (100 I.tg/ml), and 0.01% STI. All manipulations were done at 4°C. Homogenates were centrifuged at 600 rprn for 5 min in an IEC PR6000 (International Equipment Company, Needham Heights, MA) to pellet any unbroken cells. The superuatant was collected, and the pellet was rehomogenized in 1 ml of buffer using five up-down strokes with the Brendler homogenizer. Supernatant and second homogenate were combined, filtered through two layers of gauze, and aliquots were removed for DNA assays by the Burton method (4). DNA in the remaining hornogenates was then digested by incubation with 2 mg purified DNase per 4 g wet tissue weight for 60 min at 4°C. Total membranes from hornogenates containing a known quantity of DNA were then pelleted by centrifugation for 5 min at 16,000 g in an Eppendorf microfuge (Brinkrnann Instruments Co., Westbury, NY). Membranes were stored in KRH buffer under liquid N2 before binding assays. For affinity labeling of CCK binding proteins, a smooth pancreatic membrane preparation was made to reduce the amount of nonspecific labeling. Pancreata were homogenized in 10 vol/tissue wet weight of 0.3 M sucrose (containing the battery of protease inhibitors listed above) in a Dounce homogenizer using six strokes with a tight fitting pestle. The homogenate was filtered through two layers of gauze, and brought to 1.3 M sucrose by the addition of a 2-M sucrose stock solution. The volume of the 1.3 M homogenate was four times that of the 0.3-M sucrose homogenate volume (i.e., 40 times volume per tissue wet weight). The 1.3 M homogenate was loaded into centrifuge tubes, overlayed with 0.3 M sucrose, and these discontinuous The Journal of Cell Biology, Volume t03, 1986 2354 on July 1, 2017 jcb.rress.org D ow nladed fom gradients were centrifuged in a rotor (SW 41; Beckman Instruments, Inc., Fullerton, CA) for 90 min (150,000 g,0. Material banding at the 0.3-1.3-M sucrose interface was collected, diluted to a final sucrose concentration of 0.06 M with distilled H2O, and pelleted by centrifugation at 16,000 g for 5 min. The membrane pellet was resuspended in either KRH or distilled H20, stored in aliquots under liquid N2, and used immediately upon