Actions of derivatives of cyclic nucleotides on dispersed acini from guinea pig pancreas. Discovery of a competitive antagonist of the action of cholecystokinin.

In dispersed acini from guinea pig pancreas, amylase secretion was increased by BQAMP and 8Br-CAMP, but not by native cyclic AMP. These derivatives of cyclic AMP potentiated the increase in amylase secretion caused by cholecystokinin and choline&c agents but did not alter the increase in enzyme secretion caused by secretin or vasoactive intestinal peptide. Cyclic GMP did not increase amylase secretion and did not modify the increase in amylase secretion caused by various secretagogues. 8Br-cGMP increased amylase secretion apparently by virtue of its ability to mimic the action of endogenous cyclic AMP. Butyryl derivatives of cyclic GMP did not alter amylase release or the increase in amylase release caused by secretin or vasoactive intestinal peptide, but were reversible, competitive antagonists of the action of cholecystokinin and structurally related peptides. The inhibition was specific for cholecystokinin and related peptides since B&cGMP did not alter the increase in amylase secretion caused by other secretagogues (e.g. bombesin, physalaemin, carbachol, or A23187) which have a mode of action similar to that of cholecystokinin. Furthermore, this antagonism of the action of cholecystokinin depended absolutely on the presence of a butyryl moiety and the inhibitory potency of butyryl cyclic GMP depended on the number and position of the butyryl side chains.

In dispersed acini from guinea pig pancreas, amylase secretion was increased by BQAMP and 8Br-CAMP, but not by native cyclic AMP. These derivatives of cyclic AMP potentiated the increase in amylase secretion caused by cholecystokinin and choline&c agents but did not alter the increase in enzyme secretion caused by secretin or vasoactive intestinal peptide. Cyclic GMP did not increase amylase secretion and did not modify the increase in amylase secretion caused by various secretagogues. 8Br-cGMP increased amylase secretion apparently by virtue of its ability to mimic the action of endogenous cyclic AMP. Butyryl derivatives of cyclic GMP did not alter amylase release or the increase in amylase release caused by secretin or vasoactive intestinal peptide, but were reversible, competitive antagonists of the action of cholecystokinin and structurally related peptides.
The inhibition was specific for cholecystokinin and related peptides since B&cGMP did not alter the increase in amylase secretion caused by other secretagogues (e.g. bombesin, physalaemin, carbachol, or A23187) which have a mode of action similar to that of cholecystokinin. Furthermore, this antagonism of the action of cholecystokinin depended absolutely on the presence of a butyryl moiety and the inhibitory potency of butyryl cyclic GMP depended on the number and position of the butyryl side chains.

Previously
we have found that pancreatic secretagogues such as cholecystokinin and cholinergic agents act on pancreatic acinar cells to cause a significant increase in calcium outflux, cellular cyclic GMP, and amylase secretion, but do not alter cellular cyclic AMP (l-9). Secretagogues such as secretin and VIP' activate adenylate cyclase, increase cellular cyclic AMP, and stimulate amylase secretion but do not alter calcium outflux or cellular cyclic GMP (2,4,7,9). In the present studies we have examined the effects of exogenous cyclic nucleotides as well as their 8-bromo and butyryl derivatives on amylase secretion from dispersed pancreatic acini. We have also examined the abilities of exogenous cyclic nucleotides to modify the increase in amylase secretion caused by various secretagogues. We have found that amylase secretion was increased by 8Br-CAMP and B&AMP but not by native cyclic AMP. The derivatives of cyclic AMP potentiated the increase in amylase caused by cholecystokinin and cholinergic agents, but did not alter the increase caused by secretin or VIP. Two actions of the derivatives of cyclic GMP were surprising. 8Br-cGMP increased amylase secretion apparently as a result of its ability to mimic the action of endogenous cyclic AMP. Butyryl derivatives of cyclic GMP did not alter amylase secretion but were competitive antagonists of the action of cholecystokinin and structurally similar peptides. In each experiment, each value was determined in triplicate and the coefficient of variation for triplicate samples was always less than 10%.
To measure outflux of 4'Ca, acini from the pancreas of one animal were suspended in 10 ml of incubation solution containing 40 PCi of 45Ca, gassed, and preincubated at 37'C for 60 min (1,3,5,6). At the end of the preincubation the cells were washed twice by alternate centrifugation and resuspension with 100 volumes of incubation solution containing no ?a and resuspended in 40 to 60 ml of standard incubation solution. Samples (1.0 ml) of the cell suspension were incubated with the appropriate agents for 5 min at 37°C. At the beginning and at the end of the incubation acini were separated from incubation medium by centrifugation with silicone oil. A sample of cell suspension (300 ~1) was added to a microcentrifuge tube containing 106 ~1 of silicone oil and the tube was centrifuged at 10,000 x g for 60 s. One hundred microliters of supernatant was added to a counting vial containing 10 ml of Aquasol (New England Nuclear) for determination of 45Ca. One hundred microliters of cell suspension was also added directly to a counting vial for determination of radioactivity. 45Ca outflux was measured as the per cent of the total radioactivity which appeared in the incubation medium during the 5-min incubation period. In each experiment, each value was determined in triplicate and the coefficient of variation for triplicate incubations was always less than 6%. Liquid scintillation counting was performed using a Packard model 3320 liquid scintillation counter. The purity of the various cyclic nucleotide derivatives was assessed using cyclic nucleotide-specific radioimmunoassays as well as thin layer and column chromatography. Cyclic nucleotide radioimmunoassays were performed using the minor modifications of the procedure of Harper and Brooker (12) published previously (13). Thin layer chromatography was performed with preparative silica gel plates and an ascending solvent of butanol:acetic acidwater (4:l:l). The plates were illuminated with UV light and the spots visualized were scraped into 1.0 ml of distilled water and centrifuged at 1000 X g for 5 min. The supernatant was removed and boiled for 15 min to evaporate the butanol. Column chromatography was performed with Whatman DEAE-cellulose (0.6 x 6.0 cm) equilibrated with 5 mM NaCl and eluted with a linear gradient of NaCl from 5 to 300 mM.

RESULTS
Amylase release from dispersed pancreatic acini was increased by 8Br=cAMP, B&AMP and 8Br-cGMP but not by BtpcGMP ( Fig. 1) or by native cyclic nucleotides (not shown). With derivatives of cyclic AMP, the rate of amylase secretion was constant during the initial 40 min of incubation and decreased progressively thereafter ( Fig. 1). With 8Br-cGMP, the rate of amylase secretion was not altered during the initial 30 min of incubation and then increased progressively during the subsequent 90 min (Fig. 1). When present at maximally effective concentrations 8Br-CAMP caused a 7-fold increase in amylase secretion, BtzcAMP caused a 5-fold increase, and 8Br-cGMP caused a 3-fold increase (Fig. 2). 8Br-CAMP was a more potent stimulant of amylase release than was BtzcAMP and both derivatives of cyclic AMP were more potent secretagogues than 8Br-cGMP ( Fig. 2). To examine the abilities of derivatives of cyclic nucleotides to modify the increase in amylase release caused by other secretagogues, acini were preincubated with different cyclic nucleotides for 2 h at 37°C. At the end of the preincubation, the rate of amylase release was measured during a 15min incubation with no additions, CCK-OP, or secretin. Amylase release with a cyclic nucleotide derivative plus secretin was the same as that obtained with secretin alone (Table I). Similar results were obtained using VIP (10 nM) instead of secretin (not shown). Derivatives of cyclic AMP as well as 8Br-cGMP potentiated the action of CCK-OP on amylase secretion (Table I) caused by CCK-OP plus one of the cyclic nucleotide derivatives was significantly greater than the sum of the increase caused by each agent alone (Table I). Similar results were obtained using carbachol (30 pM) or bombesin (10 nM) instead of CCK-OP (not shown). In other studies, the time course for the potentiation of the action of CCK-OP by a given cyclic nucleotide derivative corresponded to the time course of the action of the same cyclic nucleotide derivative alone. Bt,cGMP, which did not alter the increase in amylase secretion caused by secretin, abolished the increase in amylase secretion caused by CCK-OP (Table I).
To examine the possibility that 8Br-cGMP was contaminated with 8Br-CAMP, we tested both cyclic nucleotide derivatives for their abilities to inhibit binding of lz51-labeled cyclic nucleotides to cyclic nucleotide-specific antibodies. Results obtained from these studies indicated that the maximal amount of 8Br-CAMP which could be present in the sample of 8Br-cGMP was 0.01%. The present results obtained with dispersed pancreatic acini differ in several respects from those obtained previously using dispersed, single acinar cells (7). Single acinar cells were prepared by incubating the tissue in a calcium-free medium containing EGTA (7), while the dispersed acini used for the present studies were prepared without using a calcium-free, EGTA-containing medium. To assess the potential influence of this calcium-free incubation on the responsiveness of the preparation to various secretagogues, dispersed acini were prepared in the usual fashion except a 20-min incubation with a calcium-free, EGTA-containing medium was added to the digestion procedure.
EGTA treatment caused a significant decrease in the stimulation of amylase secretion caused by VIP or CCK-OP (Table II). BQGMP did not alter amylase in control or EGTA-treated acini, but the nucleotide abolished the action of CCK-OP in both preparations (Table II). BtzcGMP potentiated the action of VIP in EGTA-treated acini but did not alter the action of VIP in control acini (Table  II).
In acini incubated with different concentrations of CCK-OP, amylase secretion increased, became maximal with 0.3 nM CCK-OP and then decreased as the concentration of CCK-OP was increased above 0.3 no (Fig. 3, left). Bt*cGMP caused a parallel, rightward shift in the dose-response curve for CCK-OP-stimulated amylase secretion and the magnitude of the shift was proportional to the nucleotide concentration ( Fig. 3, left). BtzcGMP did not alter the increase in amylase release caused by a maximally effective concentration of CCK-OP (Fig. 3, left). When acini were incubated with a fixed concentration of CCK-OP and different concentrations of BkGMP, two patterns of amylase secretion were observed depending on the concentration of CCK-OP. With submaximal or maximally effective concentrations of CCK-OP, BtzcGMP caused a concentration-dependent decrease in amylase secretion and with higher concentrations of CCK-OP, higher concentrations of BtzcGMP were required to abolish stimulation of amylase secretion (Fig. 3, right). With supramaximal concentrations of CCK-OP, as the concentration of BtzcGMP was increased, amylase secretion increased, became maximal, and then decreased toward basal values (Fig. 3, right). The results illustrated in Fig. 3 were obtained using BtzcGMP from Sigma Chemical Co. Similar results were obtained using Bt2cGMP from ICN or from Boehringer Mannheim Biochemicals. Results similar to those obtained measuring amylase secretion with CCK-OP and BtzcGMP were also obtained with carbachol and atropine (Fig. 4). That is, as the concentration of carbachol was increased, there was a progressive rightward shift in the dose-response curve for atropine inhibition.
Furthermore, with concentrations of carbachol which were supramaximal for amylase secretion, as the atropine concentration increased, amylase secretion increased, became maximal and then decreased to basal values (Fig. 4).
In addition to Bt2cGMP, monobutyryl derivatives of cyclic GMP also inhibited the stimulation of amylase secretion caused by CCK-OP (Fig. 5). O'-Monobutyryl cyclic GMP was approximately 20 times less potent than the dibutyryl derivative while N2-monobutyryl cyclic GMP was approximately 100 times less potent than the dibutyryl derivative (Fig. 5). In addition to native cyclic GMP, 02-tyrosine methyl ester cyclic GMP, 02'-succinyl cyclic GMP, and butyrate at concentrations as high as 10 IIIM did not alter the increase in amylase secretion caused by CCK-OP (not shown).
The inhibitory action of Bt2cGMP was specific for cholecystokinin. The nucleotide inhibited the increase in amylase secretion caused by native cholecystokinin, by COOH-terminal deca-, octa-, and heptapeptide of cholecystokinin, by analogues of the COOH-terminal heptapeptide of cholecystokinin, and by caerulein, a decapeptide in which seven of the eight COOH-terminal amino acids are identical with those in the COOH-terminal octapeptide of cholecystokinin (Table  III). BtzcGMP did not alter the increase in amylase secretion caused by bombesin, physalaemin, carbachol, A23187, secretin, or VIP (Table III) release or the ability of the nucleotide to inhibit the action of CCK-OP determined in a subsequent incubation (Table IV). In pancreatic acinar cells one of the initial effects of cholecystokinin is to cause a significant increase in calcium outflux (1,3,5,6).
In the present studies, a significant increase in *%a outflux could be detected with 0.1 nM CCK-OP and maximal stimulation occurred with 10 m CCK-OP (Fig. 6, left). Like its effect on CCK-OP-stimulated amylase secretion, Bt2cGMP caused a parallel rightward shift in the dose-response curve for the increase in 45Ca outflux caused by CCK-OP but did not alter the increase in outflux caused by a maximally effective concentration of CCK-OP (Fig. 6,  tions of Bt2cGMP were required to produce detectable inhibition of 45Ca outflux (Fig. 6, right).

30
To explore the possibility that the effects of BtzcGMP were actually due to a contaminant, the material obtained from the 25 I commercial supplier was subjected to column and to thin layer chromatography.
When chromatographed on DEAEcellulose, the nucleotide eluted as a single peak at the beginning of the salt gradient (Fig. 7). There was a close correlation between the elution profile of BtzcGMP determined by absorbance at 254 nm and that of material capable of inhibiting the increase in amylase secretion caused by 0.1 no CCK-OP (Fig. 7, inset). In each experiment, at least 87% of the applied material was eluted from the column judged by its ability to  Reversibility of action of BkcGMP on CCK-OP-stimulated amylase inhibit the increase in amylase secretion caused by 10-l" M CCK-OP.
release Amylase secretion was measured during a 30-min incubation at 37'C. Control samples obtained by eluting a column to which no cyclic Pancreatic acini were preincubated with the indicated agents for nucleotide had been applied did not alter the action of CCK-OP on 30 min at 37°C and washed twice at ambient temperature with at amylase secretion. Results shown are from one experiment and this least 100 volumes of fresh incubation solution (containing no experiment is representative of two others. B&cGMP or CCK-OP) by alternate centrifugation at 900 x g for 1 min and resuspension. The washed acini were resuspended in standard incubation solution containing the agents specified and amylase  FIG. 6. Effect of BtzcGMP on the increase in calcium outflux caused by CCK-OP. Acini were suspended in standard incubation The sample of Bt2cGMP having a RF of 0.60 caused the same solution containing 0.5 mM 45Ca and were preincubated for 60 min at inhibition of the action of CCK-OP as did an equal concentra-37°C. Acini were then washed twice with and resuspended in standard tion of stock BtzcGMP (Table V). incubation solution. Outflux of %a was measured during a 5-min incubation at 37°C with the indicated agents. In the left panel the DISCUSSION concentrations of BQGMP are given in parentheses. In the right panel the concentrations of CCK-OP are given in parentheses. Results The effects of exogenous cyclic nucleotides on pancreatic shown are means of quadruplicate determinations from one experi-enzyme secretion have not been uniform (for review see Ref. ment and this experiment is representative of four others. 14). This lack of uniformity may reflect, in part, the species of animal from which the pancreas was obtained, since in some species, increases in endogenous cyclic AMP are not accompanied by corresponding increases in enzyme secretion. For example, Robberecht et al. (15) have found that secretin and VIP can increase cyclic AMP in fragments of pancreas from dog, cat, rat, guinea pig, and mouse, but increase enzyme secretion only from guinea pig and rat pancreas. In the present studies we found that amylase secretion from dispersed pancreatic acini was increased significantly by derivatives of cyclic AMP and by 8Br-cGMP but not by native cyclic AMP, native cyclic GMP or Bt,cGMP.
With the exception of our failure to detect an increase in amylase secretion with Bt2cGMP, the present studies agree with previous studies using other preparations of guinea pig pancreas (7, 16,17).
Dispersed single acinar cells prepared by incubating the pancreas with crude collagenase, crude hyaluronidase, and EGTA differed from dispersed acini prepared with purified collagenase without hyaluronidase or EGTA in terms of their responsiveness to secretagogues as well as the pattern of action of Bt2cGMP (7). In acini the magnitude of the increase in amylase secretion caused by various secretagogues or derivatives of cyclic AMP (severalfold) is substantially greater than that observed previously in single acinar cells (less than lfold). Furthermore, unlike its action on amylase release from single acinar cells (7), in pancreatic acini BtzcGMP did not alter basal enzyme release or the increase in enzyme release caused by VIP or secretin, but inhibited the increase in enzyme release caused by CCK-OP. We do not know the basis for these differences; however, dispersed single acinar cells from mouse pancreas do not have the apical complex of microfilaments and microvilli seen in dispersed acini (18) and the loss of these structures in single acinar cells may be related to their altered responsiveness.
In addition, in the present study we found that incubating acini for 20 min with EGTA reduced the magnitude of the increase in amylase release caused by VIP or CCK-OP and modified some, but not all, of the actions of BQGMP.
These findings indicate that some of the differences between acini and single acinar cells may be attributable to the calcium-free incubation used to prepare single acinar cells.
In acini from guinea pig pancreas, amylase secretion is increased by agents which increase cellular cyclic AMP (VIP and secretin) and by agents which increase calcium outflux and cellular cyclic GMP (cholinergic agents, cholecystokinin, caerulein, bombesin, litorin, physalaemin, eledoisin, and A23187) (l-9). When a secretagogue which increases cyclic AMP is combined with a secretagogue which increases calcium outflux, there is potentiation of amylase secretion (7,9). In contrast, the increase in amylase secretion with maximally effective concentrations of two secretagogues each of which has the same mode of action is equal to that caused by the more effective secretagogue alone (7,9). In the present studies we found that 8Br-cGMP, like the derivatives of cyclic AMP, mimicked the action of agents which increase endogenous cellular cyclic AMP. Potentiation of amylase secretion occurred when 8Br-cGMP or one of the derivatives of cyclic AMP was combined with a secretagogue which increases calcium outflux and cellular cyclic GMP. The increase in amylase secretion caused by 8Br-cGMP or one of the derivatives of cyclic AMP plus a secretagogue which increases cellular cyclic AMP was the same as that with the secretagogue alone. These effects of 8Br-cGMP were not attributable to its being contaminated with 8Br-CAMP. Furthermore, the time course of action of 8Br-cGMP differed significantly from that of BBr-CAMP in that with the cyclic GMP derivative no increase in amylase secretion occurred until after 30 min of incubation.
In other tissues, exogenous derivatives of cyclic AMP have been found to cause effects similar to those caused by derivatives of cyclic GMP (for review see Ref. 19); however, in most of these studies it could not be determined whether the response was one which could be produced by both endogenous cyclic nucleotides or, as in the present study, by only one whose effect could be reproduced by exogenous derivatives of cyclic AMP and cyclic GMP. The effects of 8Br-cGMP observed in the present studies illustrate the potential for misinterpreting effects caused by exogenous derivatives of cyclic nucleotides.
Since a number of secretagogues can increase cyclic GMP in pancreatic acini and since amylase secretion is increased by 8Br-cGMP, one might conclude that the action of secretagogues which increase endogenous cyclic GMP is, in fact, mediated by cyclic GMP. The present results, however, argue against cyclic GMP being a mediator of the action of cholinergic agents or cholecystokinin since in terms of stimulating amylase secretion, 8Br-cGMP appears to exert this effect by mimicking the action of endogenous cyclic AMP. If cyclic GMP were a mediator of the action of secretagogues such as cholecystokinin and if 8Br-cGMP mimicked the action of endogenous cyclic GMP, we should have seen potentiation of amylase secretion with secretin or VIP plus 8Br-cGMP and not with CCK-OP or carbachol plus cyclic GMP.
An unanticipated finding was our observation that BtzcGMP could competitively inhibit the increase in amylase secretion caused by CCK-OP. In particular, BtzcGMP caused a parallel, rightward shift in the dose-response curve for CCK-OP-stimulated amylase release, the magnitude of this shift was proportional to the concentration of BtzcGMP and the nucleotide did not alter the maximal increase in amylase secretion caused by CCK-OP. This inhibition was fully reversible and was specific for cholecystokinin and structurally related peptides. BtzcGMP did not alter the action of secretagogues which have a mode of action similar to that of cholecystokinin and did not alter the action of secretagogues whose effects are mediated by cyclic AMP. Finally, the pattern of the effect of BtzcGMP on amylase secretion stimulated by CCK-OP was identical with that of atropine on enzyme secretion stimulated by carbachol. This ability of BttcGMP to inhibit the action of CCK-OP was not attributable to the presence of a contaminant in the material obtained from the commercial supplier, since the inhibitory activity co-chromatographed with BtzcGMP in two different systems.
Inhibition of the action of cholecystokinin by butyryl cyclic GMP required the presence of at least one butyryl moiety. Native cyclic GMP, 0"-tyrosine methyl ester cyclic GMP, and 0"-succinyl cyclic GMP did not alter the increase in amylase secretion caused by CCK-OP. Furthermore, the inhibitory potency of butyryl cyclic GMP was determined by the position and number of butyryl groups. BttcGMP was 20 times more potent than 0"-monobutyryl cyclic GMP which was 5 times more potent than N2-monobutyryl cyclic GMP. Since neither native cyclic GMP nor 8Br-cGMP was able to inhibit the action of cholecystokinin, the ability of BtzcGMP to antagonize the action of cholecystokinin does not reflect a physiologic action of native cyclic GMP but, instead reflects a pharmacologic activity which is peculiar to butyryl derivatives of cyclic GMP.
Previously, we have found that one of the earliest steps in the mechanism of action of cholecystokinin on pancreatic acinar cells is to effect a significant increase in outflux of exchangeable cellular calcium (1,3,5,6). In the present studies we found that the pattern of action of BtzcGMP on calcium outflux stimulated by CCK-OP was similar to its pattern of action on cholecystokinin-stimulated amylase secretion. Although BtzcGMP acts in a reversibly competitive fashion to inhibit one of the earliest steps in the action of cholecystokinin, we do not know whether the nucleotide derivative is functioning as a full or partial competitive antagonist (20). That is, Bt*cGMP might exert its effects by competing with cholecystokinin for occupation of the cholecystokinin receptor (full competition) or BtzcGMP might interact with sites which are functionally distinct from the cholecystokinin receptor and by so doing reduce the affinity of the cholecystokinin receptor for its ligands (partial competition).
Obviously these possibilities as well as others will require additional studies, especially, those which examine directly the interaction of cholecystokinin with its receptors on pancreatic acinar cells.

cholecystokinin. pancreas. Discovery of a competitive antagonist of the action of
Additions and Corrections Vol. 254 (1979)  Amylase release was measured during a IO-min incubation at 37°C with the indicated agents. Results are means of duplicate determinations from one experiment and this experiment is representative of two others. Duplicate values differed by less than 10%. CCK-7, CCK-8, and CCK-10 refer to the COOH-terminal heptapeptide, octapeptide, and decapeptide of cholecystokinin, respectively. CCK-33 refers to natural cholecystokinin. CCK-7(deSOa) and CCK-7(ser-SO?) refer to COOH-terminal heptapeptide of cholecystokinin with tyrosine sulfate replaced by tyrosine and serine sulfate, respectively.
Stephen L. Brenner and Edward D. Korn Page 8622, Fig. 1 Due to a printer's error, Part A of Fig. 1 was omitted. The correct figure appears below.  We suggest that subscribers photocopy these corrections and insert the photocopies at the appropriate places where the article to be corrected originally appeared. Authors are urged to introduce these corrections into any reprints they distribute. Secondary (abstract) services are urged to carry notice of these corrections as prominently as they carried the original abstracts.

12738
Additions and Corrections Vol. 254 (1979)  Amylase release was measured during a IO-min incubation at 37°C with the indicated agents. Results are means of duplicate determinations from one experiment and this experiment is representative of two others. Duplicate values differed by less than 10%. CCK-7, CCK-8, and CCK-10 refer to the COOH-terminal heptapeptide, octapeptide, and decapeptide of cholecystokinin, respectively. CCK-33 refers to natural cholecystokinin. CCK-7(deSOa) and CCK-7(ser-SO?) refer to COOH-terminal heptapeptide of cholecystokinin with tyrosine sulfate replaced by tyrosine and serine sulfate, respectively.
Stephen L. Brenner and Edward D. Korn Page 8622, Fig. 1 Due to a printer's error, Part A of Fig. 1 was omitted. The correct figure appears below.  We suggest that subscribers photocopy these corrections and insert the photocopies at the appropriate places where the article to be corrected originally appeared. Authors are urged to introduce these corrections into any reprints they distribute. Secondary (abstract) services are urged to carry notice of these corrections as prominently as they carried the original abstracts.