0021-972X/90/7005-1403$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright© 1990 by The Endocrine Society
Vol. 70, No. 5 Printed in U.S.A.
Cholecystokinin Does not Stimulate Prosomatostatin Derived Peptides in Man* RODGER A. LIDDLE AND JOHN W. ENSINCK Departments of Medicine, Mount Zion Hospital and Medical Center, University of California, San Francisco, California?94143; Duke University Medical Center and Veterans Administration Medical Center, Durham, North Carolina 27710; and University of Washington Medical Center, Seattle, Washington 98195
ABSTRACT. In man, plasma cholecystokinin (CCK) and somatostatin-28 (S-28) levels increase after ingestion of a mixed meal. Both peptides originate from the gastrointestinal tract. In supra- and periphysiological doses, CCK stimulates the release of somatostatin-14 from in vitro pancreatic islets and gastric cells and increases circulating somatostatin-like immunoreactivity in dogs, leading to the conjecture that CCK regulates somatostatin-like immunoreactivity secretion. Nonetheless, whether CCK is responsible in part for the meal-induced rise in S-28 in man has not been established. Therefore, the present study was designed to determine if CCK, at both physiological and supraphysiological concentrations, increases the circulating levels of prosomatostatin (proS)-derived peptides in humans. On 3 separate days, five healthy men ate a mixed liquid meal or received iv infusions of CCK at rates of 18 or 38 pmol/kgh. Plasma
levels of pro-S-derived peptides, including pro-S, S-14, S-13, S28, and CCK, were measured. Basal CCK levels averaged 0.9 ± 0.1 pmol/L and increased after the meal to a peak level of 5.4 ± 1.5 pmol/L and averaged 3.1 ± 1.2 pmol/L over 90 min. The mean basal levels of pro-S, S-14, and S-13, measured collectively, was 6.1 ± 0.4 pmol/L eq S14 and was unaltered by food intake. The S-28 level was 6.7 ± 0.6 pmol/L and rose to a zenith of 13.1 ± 3.3 pmol/L by 90 min. Infusion of CCK at 18 and 38 pmol/ kg-h produced steady state plasma CCK levels of 4.1 ± 1.1 and 9.9 ± 1.5 pmol/L, respectively. Basal levels of pro-S-derived peptides were unaltered during the infusion of either the low or high dose of CCK. We conclude that CCK by itself is not a physiological signal to the release of pro-S-derived peptides in man. (J Clin Endocrinol Metab 70: 1403-1407, 1990)
S
OMATOSTATIN-14 (S-14) and S-28, an NH2-terminal extension of S-14, are the two major bioactive peptides processed from the COOH-terminus of prosomatostatin (pro-S) in different cells of the brain, central nervous system, gastrointestinal (GI) tract, pancreas, thyroid, and urogenital tract (1). Both S-14, the principal product found in pancreatic and gastric D-cells and neurons, and S-28, which is synthesized in proximal intestinal epithelial cells, are secreted into the circulation (1-7). S-13, converted from S-14 by the action of tissue aminopeptidases, is also present in plasma, as is pro-S, and all of these peptides contribute to the measurement of somatostatin-like immunoreactivity (SLI), detected by RIA using antisera reacting with the COOH-terminus of pro-S (7). After the intake of a meal of mixed composition, plasma SLI levels have been repeatedly shown to rise in man and other animals (1, 6, 7). We have recently reported that this increase is due to the elevation of S-28 levels, with little or no change in pro-S, S-14, or Received September 11, 1989. Address all correspondence and requests for reprints to: Rodger A. Liddle, M.D., Department of Medicine, P.O. Box 3083, Duke University Medical Center, Durham, North Carolina 27710. * This work was supported by NIH grants DK 38626, DK 34397, and RR-37.
S-13 concentrations and concluded that most of circulating S-28 originates from the small intestine (7). Cholecystokinin (CCK) is a classical GI hormone, the actions of which include stimulation of pancreatic secretion, gallbladder contraction, regulation of gastric emptying, and bowel motility (8). There is evidence that CCK also causes the release of SLI in isolated perfused pancreas from pig and dog (9, 10). Moreover, CCK has been shown to enhance the release of SLI from rat islet cell cultures (11). When CCK-8 was infused at periphysiological doses, increases in plasma SLI levels were observed without concomitant increases in insulin or glucagon, thereby raising the possibility that one or more pro-S-related peptides originated from extrapancreatic sites (12). Since the GI tract is the major origin of circulating SLI, the stomach and/or intestine would be the most likely source of these peptides. That CCK may modulate the secretion of pro-S-related peptides is further suggested by the observation of Soil et al. (13), who found that CCK stimulated the release of S-14 from dog fundic mucosal cells in culture (13). Heretofore, difficulties in measuring plasma levels of CCK have hampered the study of the physiological effects of this hormone in man (14, 15); thus, it remains undetermined whether CCK, released during food intake,
1403
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
LIDDLE AND ENSINCK
1404
regulates the secretion of any or all of the pro-S-related peptides. Therefore, we have investigated the temporal relationships of these peptides in response to a meal using a recently developed sensitive bioassay for measurement of plasma levels of CCK combined with methods for separating the pro-S-related peptides and their measurement by RIA. To study the physiological effects of CCK on plasma levels of the pro-S-related peptides in man, we have infused CCK-8 to reproduce normal postprandial blood levels of CCK and assessed the responses of circulating pro-S-derived peptides. Materials and Methods Subjects Five healthy men, aged 24-34 yr, who were within 15% of ideal body weight, volunteered for the study. They all denied taking any medication or drugs. Written informed consent was obtained from each subject. After a 12-h overnight fast, each subject was studied in a semirecumbent position. Ingestion of liquid meal All volunteers swallowed a mixed liquid meal (400 mL) consisting of commercially available instant breakfast supplement, one egg, and "half and half milk and cream, totaling 1.5 Cal/mL (40% carbohydrate, 40% fat, and 20% protein) (14). The meal was consumed in 1-2 min, and blood was sampled from an indwelling line in a superficial vein of the forearm before and at 90-min intervals thereafter. iv infusions On separate days, each subject received infusions of 150 mM sodium chloride to which synthetic CCK-8 (Kinevac-Squibb Diagnostics, New Brunswick, NJ) was added. CCK-8 was diluted to appropriate concentrations in a total volume of 20 mL 150 mM sodium chloride. The infusion rates were determined by measuring the CCK concentration of the infusate taken from the delivery system. Blood collection and processing of plasma Fifteen milliliters of blood were collected in plastic disposable syringes and immediately placed into centrifuge tubes containing 5 nh heparin (20,000 U/mL)/mL blood. Within 5 min of drawing blood the red cells were sedimented by centrifugation. For the CCK assay the plasma was directly passed onto octadecylsilylsilica cartridges (Cis Sep-Pak, Waters Associates, Inc., Milford, MA) (14). To prevent protease-mediated destruction of pro-S-related peptides, aliquots of plasma were immediately adjusted to pH 3 by the addition of 1 N HC1 (0.1 mL/mL plasma) and stored at —20 C (7). Acidified samples were thawed and passed through Sep-Pak cartridges prepared by washing with absolute methanol and deionized water. Plasma proteins and enzymes were removed by washing sequentially with 5 mL water and 5 mL 0.1% trifluoroacetic acid in water. Adsorbed peptides were eluted with 5 mL of a solution of 80% methanol
JCE & M • 1990 Vol 70 • No 5
and 1% trifluoroacetic acid. The eluates were air dried and reconstituted in 130 mM borate buffer, pH 8.5. The recovery of authentic S-14, S-13, and S-28 in either buffer or acidified plasma was 78 ± 2% (n = 15), and these values did not differ significantly from each other. CCK bioassay CCK was adsorbed onto Sep-Pak cartridges and dried under nitrogen, and CCK bioactivity was measured as previously described, using isolated rat pancreatic acini (14, 15). The minimal detection limit was 0.2 pmol/L. Recoveries of CCK-8 and CCK-33 added to charcoal-stripped plasma were 90 ± 5% and 86 ± 12%, respectively. Intra- and interassay coefficients of variation were 7.4% and 10.4%, respectively. Plasma CCK bioactivity was reproduced by infusing synthetic CCK-8 as we have previously demonstrated in studies of gallbladder contraction and gastric emptying (14, 16). RIA of somatostatin-related peptides Details of the separation of S-28 from pro-S, S-14, and S-13 and their measurements in plasma have been previously reported (7). In brief, the samples were applied to a column of agarose coupled with partially purified immunoglobulins selectively binding the Asp5-Pro6 sequence in the NH2 region of S28. The column was washed with 130 mM borate buffer, pH 8.5, in which S-14, S-13, and pro-S were eluted. S-28, adherent to the immunoadsorbent, was removed with 0.2 N acetic acid containing 0.2% BSA, pH 3.5, and lyophilized. The samples in the fall-through fraction (pro-S, S-14, and S-13) and S-28, eluted from the immunoadsorbent and reconstituted in 130 mM borate buffer, were assayed using antiserum AS-10, which recognizes pro-S, S-14, S-13, and S-28. Antiserum AS-10 interacts with the Phe7-Trp8-Lys9 residues of S-14. From stoichiometric comparisons, S-14 and S-13 bound equivalently with AS-10, whereas S-28 had only one third of the avidity of S-14, implying that the NH2-terminal extension in S-28 partially occludes ligand-antibody interaction. Precise quantification of pro-S was not possible because of limited amounts of this peptide. Assays were carried out in 130 mM borate buffer, pH 8.5, containing 0.2% BSA, with AS-10 antiserum diluted to 1:100,000 using [125I]Tyrn-S-14 as tracer. Free and antibodybound labeled peptides were separated by the addition of 1 mL 1% activated charcoal (Norit-A, Eastman Kodak, Rochester, NY). The values for pro-S, S-14, and S-13 collectively were read against S-14 as standard and expressed as picomoles per L eq. Measurements were corrected for an average recovery of 78%. The values for S-28 were expressed as picomoles per L, using synthetic S-28 as standard. Measurements of S-28 were corrected for an average recovery of 50% when passed through both Sep-Pak cartridges and the immunoadsorbent. Statistical analysis All values are expressed as the mean ± SEM. Differences in plasma peptide levels were analyzed using the Wilcoxon nonparametric signed rank test, comparing the areas under the curve. P < 0.05 was considered significant.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
CCK AND SOMATOSTATIN SECRETION meal
Results 8^
Responses of CCK and somatostatin-related peptides to a liquid meal Plasma levels of CCK and pro-S-derived peptides before and after the intake of a mixed meal are shown in Figs. 1 and 2. Basal CCK levels averaged 0.9 ± 0.1 pmol/ L and increased promptly to a peak level of 5.4 ± 1.5 pmol/L by 20 min after the meal. Levels then gradually declined thereafter, but remained significantly above basal levels for 90 min. Plasma levels of pro-S, S-14, and S-13 averaged 6.1 ± 0.4 pmol/Leq S-14 at baseline and did not change significantly after the meal (Fig. 2A). In contrast, S-28 levels rose promptly by 30 min and achieved a peak level of 13.1 ± 3.3 pmol/L by 90 min postprandially.
•
^"i
*L cr to a>
end 60
6 4-
5 -10 0
20 i
To determine if the elevated plasma CCK levels that occurred after food intake might contribute to the increase in plasma S-28 levels, S-28 concentrations were measured after the infusion of CCK-8. Two doses of CCK were chosen to span the physiological range of plasma CCK levels (Fig. 3). At a CCK dose of 18 pmol/ kg-h, steady state plasma CCK levels averaged 4.1 ± 1.1 pmol/L, which corresponds to the mean level obtained postprandially (Fig. 1). When CCK was infused at a rate of 38 pmol/kgh, a plasma level of 9.9 ± 1.5 pmol/L was
10-
5 -
90
1 0
60
90
meal
3
-10 0
30
Minutes FIG. 2. Plasma levels of pro-S-derived peptides in response to a meal. The subjects described in Fig. 1 consumed a liquid meal (600 Cal), and levels of pro-S, S-14, and S-13 (A) and S-28 (B) were determined at intervals thereafter. Results for pro-S, S-14, and S-13 in combination are expressed as picomoles per L eq S-14, and those for S-28 as picomoles per L. The symbols represent the mean ± SEM. ®, P < 0.05 compared with basal values.
2 -
achieved, representing an approximate doubling from the level in the postcibum period. The levels of pro-S-related peptides after infusion of CCK are shown in Fig. 4. Despite attaining physiological levels of CCK at the lower rate of CCK infusion, pro-S, S-14, and S-13 levels were unaltered (Fig. 4A), as were levels of S-28 (Fig. 4B). Similarly, when supraphysiological levels of CCK were achieved, none of the pro-S-derived peptide levels were significantly changed from baseline values.
1 -
Discussion
4 O
3 -
o o
60
B
15-
meal
30 Minutes
Responses of somatostatin-derived peptides to CCK-8 infusion
6 -i
1405
0J
-10 0
30
60
90
Minutes FIG. 1. Plasma levels of CCK in response to a meal. Five healthy men ingested a mixed meal (600 Cal), and plasma CCK was measured at the times indicated. Results are expressed as the mean ± SEM. ®, P < 0.05 compared with basal values.
The present study was designed to determine if CCK has a stimulatory effect on the secretion of peptides derived from pro-S in humans. We have shown that there is a temporal relationship between the release of CCK and S-28 after the intake of a mixed meal, with CCK achieving its peak response in plasma earlier than S-28, both of which are secreted from the small intestine. This differential release between these two peptides, in previous in vitro studies and in animals, indicating that
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
LIDDLE AND ENSINCK
1406
JCE & M • 1990 Vol 70 • No 5
CCK Infusion
CCK Infusion 25-1
12 n 20 -
10 15 -
8 -
o E
6 -
o o
4 -
10 -
CCK (18 pmol/kgh)
21 5 -
0 -
2 -
-10 0
CCK (18pmol/kgh)
30
60
90
Minutes
0
J
CCK Infusion
-10 0
30
60
90
B 151
Minutes FIG. 3. Plasma CCK levels during infusion of CCK-8. On separate days subjects received iv infusions of CCK-8 at 18 and 38 pmol/kgh, respectively. Plasma CCK levels achieved at all time points after baseline were significantly greater than basal values (P < 0.05).
CCK (18 pmol/kgh)
10 o
pharmacological and periphysiological doses of CCK evoke SLI release, might be construed as a link between a possible physiological effect of CCK on secretion of one or more pro-S-derived peptides (9-12, 17, 18). Schusdziarra et al. (12) reported that plasma SLI levels were increased by infusion of CCK-8 at 10-30 pmol/gh, and the increments were not further augmented by coinfusion of either glucose or amino acids. It should be noted, however, that no measurements of plasma CCK levels were obtained. In our hands, the infusions of CCK8 that achieved physiological and supraphysiological plasma concentrations of CCK in man failed to alter the circulating levels of any of the bioactive peptides cleaved from the COOH-terminus of pro-S. Although we did not attempt to mimick the rise in circulating substrates that occurs postprandially by infusing them parenterally, the finding of Schusdziarra and colleagues (12) that no additive effect in dogs occurred under such circumstances makes it unlikely that we would have found potentiation of a CCK effect on S-28 release by nutrients in man. Therefore, we conclude that the rise in S-28 levels after a meal in man is not due to a direct effect of CCK. Whether other enteral peptides secreted during absorption of food might stimulate the release of S-28 or S-14 under physiological conditions has not been critically examined. Rouiller et al. (19) have compared the effects of infusion of CCK-8, secretin, and gastric inhibitory polypeptide in dogs with measurement of SLI in
a. CCK (38 pmol/kgh)
CO
5-
0J
-10 0
30
60
90
Minutes FIG. 4. Plasma levels of pro-S, S-14, and S-13 in combination and of S-28 during infusions of two doses of CCK. Refer to Fig. 3 for details.
venous drainage from stomach, pancreas, and inferior vena cava. They found that CCK-8 was the most potent in releasing SLI from pancreas and stomach. Changes in SLI levels in the inferior vena cava, however, were less than one third those that would have been seen with a large fat-protein meal. Although species differences may account for our inability to confirm rises of somatostatinrelated peptides in man after CCK-8 administration, the rank order of effect of the putative pharmacological levels of peptide hormones achieved in the dog (CCK-8 > secretin > gastric inhibitory polypeptide) (19) would make it even less likely that any of the other peptides would have a significant effect on S-28 secretion in man. We have found that fat and, to a lesser extent, protein, but not carbohydrate, in food lead to enhanced S-28 secretion in humans (20). Mounting evidence supports the possibility that the nutrient-induced increase in S-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
CCK AND SOMATOSTATIN SECRETION
28 may be mediated by the cholinergic nervous system, which is stimulated through receptors sensitive to lipids and amino acids (21-24). From our present studies we contend that in man neither CCK nor other enteropeptides plays a significant role in S-28 release. In summary, we have addressed the question of whether CCK may be a modulator of S-28 release which is increased during nutrient intake. Although we have shown a temporal relationship between the secretion of CCK and S-28 after food ingestion, the infusion of CCK8 to mimic periphysiological and supraphysiological levels of CCK did not increase levels of any of the pro-Sderived peptides. We conclude that CCK is not a significant physiological stimulus to S-28 secretion.
Acknowledgments The authors thank Ms. Jacqueline Carter and Ms. Robin Vogel for excellent technical assistance.
References 1. Reichlin S. Somatostatin. N Engl J Med. 1983;309:1495-501,155663. 2. Robbins RJ, Reichlin S. Somatostatin biosynthesis by cerebral cortical cells in monolayer culture. Endocrinology. 1983; 113:57481. 3. Patel YC. A high molecular weight form of somatostatin-28 (1-12like) immunoreactive substance without somatostatin-14 immunoreactivity in the rat pancreas. J Clin Invest. 1983;72:2137-43. 4. Chiba T, Park J, Yamada T. Biosynthesis of somatostatin in canine fundic D-cells. J Clin Invest. 1988;81:282-7. 5. Baskin DC, Ensinck JW. Somatostatin in epithelial cells of intestinal mucosa is present primarily as somatostatin-28. Peptides. 1984;5:615-21. 6. Shoelson SE, Polonsky KS, Nakabayashi T, Jaspan JB, Tager HS. Circulating forms of somatostatinlike immunoreactivity in human plasma. Am J Physiol. 1989;250:E428-34. 7. Ensinck JW, Laschansky EC, Vogel RE, Simonowitz DA, Roos BA, Francis BH. Circulating prosomatostatin-derived peptides: differential responses to food ingestion. J Clin Invest. 1989;83:1580-9. 8. Mutt V. Secretin and cholecystokinin. In: Mutt V, ed. Gastrointestinal hormones. San Diego: Academic Press; 1988;251-320. 9. Rehfeld JF, Larsson LI, Goltermann NR, et al. Neural regulation of pancreatic hormone secretion by the C-terminal tetrapeptide of CCK. Nature. 1980;284:33-8.
1407
10. Hermansen K. Effects of cholecystokinin (CCK)-4, nonsulfated CCK-8, and sulfated CCK-8 on pancreatic somatostatin, insulin, and glucagon secretion in the dog: studies in vitro. Endocrinology. 1984; 114:1770-5. 11. Fedotov VP, Sadovnikova NV, Gudoshnikov VI, et al. Effect of Cterminal tetrapeptide cholecystokinin (CCK-4) on function of the islands of Langerhans and the adenohypophysis. Bull Exp Biol Med. 1984;97:729-31. 12. Schusdziarra V, Lenz N, Schick R, Maier V. Modulatory effect of glucose, amino acids, and secretin on CCK-8-induced somatostatin and pancreatic polypeptide release in dogs. Diabetes. 1986;35:5239. 13. Soil AH, Amirian DA, Park J, Elashoff JD, Yamada T. Cholecystokinin potently releases somatostatin from canine fundic mucosal cells in short-term culture. Am J Physiol. 1985;248:G569-73. 14. Liddle RA, Goldfine ID, Rosen MS, Taplitz RA, Williams JA. Cholecystokinin bioactivity in human plasma: molecular forms, responses to feeding, and relationship to gallbladder contraction. J Clin Invest. 1985;75:1144-52. 15. Liddle RA, Goldfine ID, Williams JA. Bioassay of plasma cholecystokinin in rats: effects of food, trypsin inhibitor, and alcohol. Gastroenterology. 1984;87:542-9. 16. Liddle RA, Morita ET, Conrad CK, Williams JA. Regulation of gastric emptying in humans by cholecystokinin. J Clin Invest. 1986;77:992-6. 17. Ipp E, Dobbs RE, Harris V, Arimura A, Vale W, Unger RH. The effects of gastrin and gastric-inhibitory polypeptide, secretin and octapeptide of cholecystokinin upon immunoreactive somatostatin release by the perfused canine pancreas. J Clin Invest. 1977;60:1216-9. 18. Ipp E, Dobbs RE, Arimura A, Vale W, Harris V, Unger RH. Release of immunoreactive somatostatin from the pancreas in response to glucose, amino acids, pancreozymin-cholecystokinin and tobutamide. Clin Invest. 1977;60:760-5. 19. Rouiller D, Schusdziarra V, Harris V, Unger RH. Release of pancreatic and gastric somatostatin-like immunoreactivity in response to the octapeptide of cholecystokinin, secretin, gastric inhibitory polypeptide and gastrin-17 in dogs. Endocrinology. 1980;107:5249. 20. Ensinck JW, Vogel RE, Laschansky EC, Francis BH. Effect of ingested carbohydrate, fat, and protein on the release of somatostatin 28 in man. Gastroenterology. 1990;98:633-8. 21. Lucey MR, Wass JAH, Fairclough P, Webb S, Medbak S, Reese LH. Autonomic regulation of postprandial plasma somatostatin, gastrin, and insulin. Gut. 1985;26:683-8. 22. Wussen-Colle M-C, Lalieu C, Simoens C, DeGraef J. Effect of vagotomy and atropine on plasma somatostatin response to a meal in conscious dogs. Regul Peptides. 1988;21:29-36. 23. Melone J. Vagal receptors sensitive to lipids in small intestine of the cat. J Autonom Nerv Syst. 1986;17:231-41. 24. Jeanningros R. Vagal unitary response to intestinal amino acid infusions in the anesthetized cat: a putative signal for proteininduced satiety. Physiol Behav. 1982;28:9-21.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
Vol. 70, No. 5 Printed in U.S.A.
Cholecystokinin Does not Stimulate Prosomatostatin Derived Peptides in Man* RODGER A. LIDDLE AND JOHN W. ENSINCK Departments of Medicine, Mount Zion Hospital and Medical Center, University of California, San Francisco, California?94143; Duke University Medical Center and Veterans Administration Medical Center, Durham, North Carolina 27710; and University of Washington Medical Center, Seattle, Washington 98195
ABSTRACT. In man, plasma cholecystokinin (CCK) and somatostatin-28 (S-28) levels increase after ingestion of a mixed meal. Both peptides originate from the gastrointestinal tract. In supra- and periphysiological doses, CCK stimulates the release of somatostatin-14 from in vitro pancreatic islets and gastric cells and increases circulating somatostatin-like immunoreactivity in dogs, leading to the conjecture that CCK regulates somatostatin-like immunoreactivity secretion. Nonetheless, whether CCK is responsible in part for the meal-induced rise in S-28 in man has not been established. Therefore, the present study was designed to determine if CCK, at both physiological and supraphysiological concentrations, increases the circulating levels of prosomatostatin (proS)-derived peptides in humans. On 3 separate days, five healthy men ate a mixed liquid meal or received iv infusions of CCK at rates of 18 or 38 pmol/kgh. Plasma
levels of pro-S-derived peptides, including pro-S, S-14, S-13, S28, and CCK, were measured. Basal CCK levels averaged 0.9 ± 0.1 pmol/L and increased after the meal to a peak level of 5.4 ± 1.5 pmol/L and averaged 3.1 ± 1.2 pmol/L over 90 min. The mean basal levels of pro-S, S-14, and S-13, measured collectively, was 6.1 ± 0.4 pmol/L eq S14 and was unaltered by food intake. The S-28 level was 6.7 ± 0.6 pmol/L and rose to a zenith of 13.1 ± 3.3 pmol/L by 90 min. Infusion of CCK at 18 and 38 pmol/ kg-h produced steady state plasma CCK levels of 4.1 ± 1.1 and 9.9 ± 1.5 pmol/L, respectively. Basal levels of pro-S-derived peptides were unaltered during the infusion of either the low or high dose of CCK. We conclude that CCK by itself is not a physiological signal to the release of pro-S-derived peptides in man. (J Clin Endocrinol Metab 70: 1403-1407, 1990)
S
OMATOSTATIN-14 (S-14) and S-28, an NH2-terminal extension of S-14, are the two major bioactive peptides processed from the COOH-terminus of prosomatostatin (pro-S) in different cells of the brain, central nervous system, gastrointestinal (GI) tract, pancreas, thyroid, and urogenital tract (1). Both S-14, the principal product found in pancreatic and gastric D-cells and neurons, and S-28, which is synthesized in proximal intestinal epithelial cells, are secreted into the circulation (1-7). S-13, converted from S-14 by the action of tissue aminopeptidases, is also present in plasma, as is pro-S, and all of these peptides contribute to the measurement of somatostatin-like immunoreactivity (SLI), detected by RIA using antisera reacting with the COOH-terminus of pro-S (7). After the intake of a meal of mixed composition, plasma SLI levels have been repeatedly shown to rise in man and other animals (1, 6, 7). We have recently reported that this increase is due to the elevation of S-28 levels, with little or no change in pro-S, S-14, or Received September 11, 1989. Address all correspondence and requests for reprints to: Rodger A. Liddle, M.D., Department of Medicine, P.O. Box 3083, Duke University Medical Center, Durham, North Carolina 27710. * This work was supported by NIH grants DK 38626, DK 34397, and RR-37.
S-13 concentrations and concluded that most of circulating S-28 originates from the small intestine (7). Cholecystokinin (CCK) is a classical GI hormone, the actions of which include stimulation of pancreatic secretion, gallbladder contraction, regulation of gastric emptying, and bowel motility (8). There is evidence that CCK also causes the release of SLI in isolated perfused pancreas from pig and dog (9, 10). Moreover, CCK has been shown to enhance the release of SLI from rat islet cell cultures (11). When CCK-8 was infused at periphysiological doses, increases in plasma SLI levels were observed without concomitant increases in insulin or glucagon, thereby raising the possibility that one or more pro-S-related peptides originated from extrapancreatic sites (12). Since the GI tract is the major origin of circulating SLI, the stomach and/or intestine would be the most likely source of these peptides. That CCK may modulate the secretion of pro-S-related peptides is further suggested by the observation of Soil et al. (13), who found that CCK stimulated the release of S-14 from dog fundic mucosal cells in culture (13). Heretofore, difficulties in measuring plasma levels of CCK have hampered the study of the physiological effects of this hormone in man (14, 15); thus, it remains undetermined whether CCK, released during food intake,
1403
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
LIDDLE AND ENSINCK
1404
regulates the secretion of any or all of the pro-S-related peptides. Therefore, we have investigated the temporal relationships of these peptides in response to a meal using a recently developed sensitive bioassay for measurement of plasma levels of CCK combined with methods for separating the pro-S-related peptides and their measurement by RIA. To study the physiological effects of CCK on plasma levels of the pro-S-related peptides in man, we have infused CCK-8 to reproduce normal postprandial blood levels of CCK and assessed the responses of circulating pro-S-derived peptides. Materials and Methods Subjects Five healthy men, aged 24-34 yr, who were within 15% of ideal body weight, volunteered for the study. They all denied taking any medication or drugs. Written informed consent was obtained from each subject. After a 12-h overnight fast, each subject was studied in a semirecumbent position. Ingestion of liquid meal All volunteers swallowed a mixed liquid meal (400 mL) consisting of commercially available instant breakfast supplement, one egg, and "half and half milk and cream, totaling 1.5 Cal/mL (40% carbohydrate, 40% fat, and 20% protein) (14). The meal was consumed in 1-2 min, and blood was sampled from an indwelling line in a superficial vein of the forearm before and at 90-min intervals thereafter. iv infusions On separate days, each subject received infusions of 150 mM sodium chloride to which synthetic CCK-8 (Kinevac-Squibb Diagnostics, New Brunswick, NJ) was added. CCK-8 was diluted to appropriate concentrations in a total volume of 20 mL 150 mM sodium chloride. The infusion rates were determined by measuring the CCK concentration of the infusate taken from the delivery system. Blood collection and processing of plasma Fifteen milliliters of blood were collected in plastic disposable syringes and immediately placed into centrifuge tubes containing 5 nh heparin (20,000 U/mL)/mL blood. Within 5 min of drawing blood the red cells were sedimented by centrifugation. For the CCK assay the plasma was directly passed onto octadecylsilylsilica cartridges (Cis Sep-Pak, Waters Associates, Inc., Milford, MA) (14). To prevent protease-mediated destruction of pro-S-related peptides, aliquots of plasma were immediately adjusted to pH 3 by the addition of 1 N HC1 (0.1 mL/mL plasma) and stored at —20 C (7). Acidified samples were thawed and passed through Sep-Pak cartridges prepared by washing with absolute methanol and deionized water. Plasma proteins and enzymes were removed by washing sequentially with 5 mL water and 5 mL 0.1% trifluoroacetic acid in water. Adsorbed peptides were eluted with 5 mL of a solution of 80% methanol
JCE & M • 1990 Vol 70 • No 5
and 1% trifluoroacetic acid. The eluates were air dried and reconstituted in 130 mM borate buffer, pH 8.5. The recovery of authentic S-14, S-13, and S-28 in either buffer or acidified plasma was 78 ± 2% (n = 15), and these values did not differ significantly from each other. CCK bioassay CCK was adsorbed onto Sep-Pak cartridges and dried under nitrogen, and CCK bioactivity was measured as previously described, using isolated rat pancreatic acini (14, 15). The minimal detection limit was 0.2 pmol/L. Recoveries of CCK-8 and CCK-33 added to charcoal-stripped plasma were 90 ± 5% and 86 ± 12%, respectively. Intra- and interassay coefficients of variation were 7.4% and 10.4%, respectively. Plasma CCK bioactivity was reproduced by infusing synthetic CCK-8 as we have previously demonstrated in studies of gallbladder contraction and gastric emptying (14, 16). RIA of somatostatin-related peptides Details of the separation of S-28 from pro-S, S-14, and S-13 and their measurements in plasma have been previously reported (7). In brief, the samples were applied to a column of agarose coupled with partially purified immunoglobulins selectively binding the Asp5-Pro6 sequence in the NH2 region of S28. The column was washed with 130 mM borate buffer, pH 8.5, in which S-14, S-13, and pro-S were eluted. S-28, adherent to the immunoadsorbent, was removed with 0.2 N acetic acid containing 0.2% BSA, pH 3.5, and lyophilized. The samples in the fall-through fraction (pro-S, S-14, and S-13) and S-28, eluted from the immunoadsorbent and reconstituted in 130 mM borate buffer, were assayed using antiserum AS-10, which recognizes pro-S, S-14, S-13, and S-28. Antiserum AS-10 interacts with the Phe7-Trp8-Lys9 residues of S-14. From stoichiometric comparisons, S-14 and S-13 bound equivalently with AS-10, whereas S-28 had only one third of the avidity of S-14, implying that the NH2-terminal extension in S-28 partially occludes ligand-antibody interaction. Precise quantification of pro-S was not possible because of limited amounts of this peptide. Assays were carried out in 130 mM borate buffer, pH 8.5, containing 0.2% BSA, with AS-10 antiserum diluted to 1:100,000 using [125I]Tyrn-S-14 as tracer. Free and antibodybound labeled peptides were separated by the addition of 1 mL 1% activated charcoal (Norit-A, Eastman Kodak, Rochester, NY). The values for pro-S, S-14, and S-13 collectively were read against S-14 as standard and expressed as picomoles per L eq. Measurements were corrected for an average recovery of 78%. The values for S-28 were expressed as picomoles per L, using synthetic S-28 as standard. Measurements of S-28 were corrected for an average recovery of 50% when passed through both Sep-Pak cartridges and the immunoadsorbent. Statistical analysis All values are expressed as the mean ± SEM. Differences in plasma peptide levels were analyzed using the Wilcoxon nonparametric signed rank test, comparing the areas under the curve. P < 0.05 was considered significant.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
CCK AND SOMATOSTATIN SECRETION meal
Results 8^
Responses of CCK and somatostatin-related peptides to a liquid meal Plasma levels of CCK and pro-S-derived peptides before and after the intake of a mixed meal are shown in Figs. 1 and 2. Basal CCK levels averaged 0.9 ± 0.1 pmol/ L and increased promptly to a peak level of 5.4 ± 1.5 pmol/L by 20 min after the meal. Levels then gradually declined thereafter, but remained significantly above basal levels for 90 min. Plasma levels of pro-S, S-14, and S-13 averaged 6.1 ± 0.4 pmol/Leq S-14 at baseline and did not change significantly after the meal (Fig. 2A). In contrast, S-28 levels rose promptly by 30 min and achieved a peak level of 13.1 ± 3.3 pmol/L by 90 min postprandially.
•
^"i
*L cr to a>
end 60
6 4-
5 -10 0
20 i
To determine if the elevated plasma CCK levels that occurred after food intake might contribute to the increase in plasma S-28 levels, S-28 concentrations were measured after the infusion of CCK-8. Two doses of CCK were chosen to span the physiological range of plasma CCK levels (Fig. 3). At a CCK dose of 18 pmol/ kg-h, steady state plasma CCK levels averaged 4.1 ± 1.1 pmol/L, which corresponds to the mean level obtained postprandially (Fig. 1). When CCK was infused at a rate of 38 pmol/kgh, a plasma level of 9.9 ± 1.5 pmol/L was
10-
5 -
90
1 0
60
90
meal
3
-10 0
30
Minutes FIG. 2. Plasma levels of pro-S-derived peptides in response to a meal. The subjects described in Fig. 1 consumed a liquid meal (600 Cal), and levels of pro-S, S-14, and S-13 (A) and S-28 (B) were determined at intervals thereafter. Results for pro-S, S-14, and S-13 in combination are expressed as picomoles per L eq S-14, and those for S-28 as picomoles per L. The symbols represent the mean ± SEM. ®, P < 0.05 compared with basal values.
2 -
achieved, representing an approximate doubling from the level in the postcibum period. The levels of pro-S-related peptides after infusion of CCK are shown in Fig. 4. Despite attaining physiological levels of CCK at the lower rate of CCK infusion, pro-S, S-14, and S-13 levels were unaltered (Fig. 4A), as were levels of S-28 (Fig. 4B). Similarly, when supraphysiological levels of CCK were achieved, none of the pro-S-derived peptide levels were significantly changed from baseline values.
1 -
Discussion
4 O
3 -
o o
60
B
15-
meal
30 Minutes
Responses of somatostatin-derived peptides to CCK-8 infusion
6 -i
1405
0J
-10 0
30
60
90
Minutes FIG. 1. Plasma levels of CCK in response to a meal. Five healthy men ingested a mixed meal (600 Cal), and plasma CCK was measured at the times indicated. Results are expressed as the mean ± SEM. ®, P < 0.05 compared with basal values.
The present study was designed to determine if CCK has a stimulatory effect on the secretion of peptides derived from pro-S in humans. We have shown that there is a temporal relationship between the release of CCK and S-28 after the intake of a mixed meal, with CCK achieving its peak response in plasma earlier than S-28, both of which are secreted from the small intestine. This differential release between these two peptides, in previous in vitro studies and in animals, indicating that
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
LIDDLE AND ENSINCK
1406
JCE & M • 1990 Vol 70 • No 5
CCK Infusion
CCK Infusion 25-1
12 n 20 -
10 15 -
8 -
o E
6 -
o o
4 -
10 -
CCK (18 pmol/kgh)
21 5 -
0 -
2 -
-10 0
CCK (18pmol/kgh)
30
60
90
Minutes
0
J
CCK Infusion
-10 0
30
60
90
B 151
Minutes FIG. 3. Plasma CCK levels during infusion of CCK-8. On separate days subjects received iv infusions of CCK-8 at 18 and 38 pmol/kgh, respectively. Plasma CCK levels achieved at all time points after baseline were significantly greater than basal values (P < 0.05).
CCK (18 pmol/kgh)
10 o
pharmacological and periphysiological doses of CCK evoke SLI release, might be construed as a link between a possible physiological effect of CCK on secretion of one or more pro-S-derived peptides (9-12, 17, 18). Schusdziarra et al. (12) reported that plasma SLI levels were increased by infusion of CCK-8 at 10-30 pmol/gh, and the increments were not further augmented by coinfusion of either glucose or amino acids. It should be noted, however, that no measurements of plasma CCK levels were obtained. In our hands, the infusions of CCK8 that achieved physiological and supraphysiological plasma concentrations of CCK in man failed to alter the circulating levels of any of the bioactive peptides cleaved from the COOH-terminus of pro-S. Although we did not attempt to mimick the rise in circulating substrates that occurs postprandially by infusing them parenterally, the finding of Schusdziarra and colleagues (12) that no additive effect in dogs occurred under such circumstances makes it unlikely that we would have found potentiation of a CCK effect on S-28 release by nutrients in man. Therefore, we conclude that the rise in S-28 levels after a meal in man is not due to a direct effect of CCK. Whether other enteral peptides secreted during absorption of food might stimulate the release of S-28 or S-14 under physiological conditions has not been critically examined. Rouiller et al. (19) have compared the effects of infusion of CCK-8, secretin, and gastric inhibitory polypeptide in dogs with measurement of SLI in
a. CCK (38 pmol/kgh)
CO
5-
0J
-10 0
30
60
90
Minutes FIG. 4. Plasma levels of pro-S, S-14, and S-13 in combination and of S-28 during infusions of two doses of CCK. Refer to Fig. 3 for details.
venous drainage from stomach, pancreas, and inferior vena cava. They found that CCK-8 was the most potent in releasing SLI from pancreas and stomach. Changes in SLI levels in the inferior vena cava, however, were less than one third those that would have been seen with a large fat-protein meal. Although species differences may account for our inability to confirm rises of somatostatinrelated peptides in man after CCK-8 administration, the rank order of effect of the putative pharmacological levels of peptide hormones achieved in the dog (CCK-8 > secretin > gastric inhibitory polypeptide) (19) would make it even less likely that any of the other peptides would have a significant effect on S-28 secretion in man. We have found that fat and, to a lesser extent, protein, but not carbohydrate, in food lead to enhanced S-28 secretion in humans (20). Mounting evidence supports the possibility that the nutrient-induced increase in S-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
CCK AND SOMATOSTATIN SECRETION
28 may be mediated by the cholinergic nervous system, which is stimulated through receptors sensitive to lipids and amino acids (21-24). From our present studies we contend that in man neither CCK nor other enteropeptides plays a significant role in S-28 release. In summary, we have addressed the question of whether CCK may be a modulator of S-28 release which is increased during nutrient intake. Although we have shown a temporal relationship between the secretion of CCK and S-28 after food ingestion, the infusion of CCK8 to mimic periphysiological and supraphysiological levels of CCK did not increase levels of any of the pro-Sderived peptides. We conclude that CCK is not a significant physiological stimulus to S-28 secretion.
Acknowledgments The authors thank Ms. Jacqueline Carter and Ms. Robin Vogel for excellent technical assistance.
References 1. Reichlin S. Somatostatin. N Engl J Med. 1983;309:1495-501,155663. 2. Robbins RJ, Reichlin S. Somatostatin biosynthesis by cerebral cortical cells in monolayer culture. Endocrinology. 1983; 113:57481. 3. Patel YC. A high molecular weight form of somatostatin-28 (1-12like) immunoreactive substance without somatostatin-14 immunoreactivity in the rat pancreas. J Clin Invest. 1983;72:2137-43. 4. Chiba T, Park J, Yamada T. Biosynthesis of somatostatin in canine fundic D-cells. J Clin Invest. 1988;81:282-7. 5. Baskin DC, Ensinck JW. Somatostatin in epithelial cells of intestinal mucosa is present primarily as somatostatin-28. Peptides. 1984;5:615-21. 6. Shoelson SE, Polonsky KS, Nakabayashi T, Jaspan JB, Tager HS. Circulating forms of somatostatinlike immunoreactivity in human plasma. Am J Physiol. 1989;250:E428-34. 7. Ensinck JW, Laschansky EC, Vogel RE, Simonowitz DA, Roos BA, Francis BH. Circulating prosomatostatin-derived peptides: differential responses to food ingestion. J Clin Invest. 1989;83:1580-9. 8. Mutt V. Secretin and cholecystokinin. In: Mutt V, ed. Gastrointestinal hormones. San Diego: Academic Press; 1988;251-320. 9. Rehfeld JF, Larsson LI, Goltermann NR, et al. Neural regulation of pancreatic hormone secretion by the C-terminal tetrapeptide of CCK. Nature. 1980;284:33-8.
1407
10. Hermansen K. Effects of cholecystokinin (CCK)-4, nonsulfated CCK-8, and sulfated CCK-8 on pancreatic somatostatin, insulin, and glucagon secretion in the dog: studies in vitro. Endocrinology. 1984; 114:1770-5. 11. Fedotov VP, Sadovnikova NV, Gudoshnikov VI, et al. Effect of Cterminal tetrapeptide cholecystokinin (CCK-4) on function of the islands of Langerhans and the adenohypophysis. Bull Exp Biol Med. 1984;97:729-31. 12. Schusdziarra V, Lenz N, Schick R, Maier V. Modulatory effect of glucose, amino acids, and secretin on CCK-8-induced somatostatin and pancreatic polypeptide release in dogs. Diabetes. 1986;35:5239. 13. Soil AH, Amirian DA, Park J, Elashoff JD, Yamada T. Cholecystokinin potently releases somatostatin from canine fundic mucosal cells in short-term culture. Am J Physiol. 1985;248:G569-73. 14. Liddle RA, Goldfine ID, Rosen MS, Taplitz RA, Williams JA. Cholecystokinin bioactivity in human plasma: molecular forms, responses to feeding, and relationship to gallbladder contraction. J Clin Invest. 1985;75:1144-52. 15. Liddle RA, Goldfine ID, Williams JA. Bioassay of plasma cholecystokinin in rats: effects of food, trypsin inhibitor, and alcohol. Gastroenterology. 1984;87:542-9. 16. Liddle RA, Morita ET, Conrad CK, Williams JA. Regulation of gastric emptying in humans by cholecystokinin. J Clin Invest. 1986;77:992-6. 17. Ipp E, Dobbs RE, Harris V, Arimura A, Vale W, Unger RH. The effects of gastrin and gastric-inhibitory polypeptide, secretin and octapeptide of cholecystokinin upon immunoreactive somatostatin release by the perfused canine pancreas. J Clin Invest. 1977;60:1216-9. 18. Ipp E, Dobbs RE, Arimura A, Vale W, Harris V, Unger RH. Release of immunoreactive somatostatin from the pancreas in response to glucose, amino acids, pancreozymin-cholecystokinin and tobutamide. Clin Invest. 1977;60:760-5. 19. Rouiller D, Schusdziarra V, Harris V, Unger RH. Release of pancreatic and gastric somatostatin-like immunoreactivity in response to the octapeptide of cholecystokinin, secretin, gastric inhibitory polypeptide and gastrin-17 in dogs. Endocrinology. 1980;107:5249. 20. Ensinck JW, Vogel RE, Laschansky EC, Francis BH. Effect of ingested carbohydrate, fat, and protein on the release of somatostatin 28 in man. Gastroenterology. 1990;98:633-8. 21. Lucey MR, Wass JAH, Fairclough P, Webb S, Medbak S, Reese LH. Autonomic regulation of postprandial plasma somatostatin, gastrin, and insulin. Gut. 1985;26:683-8. 22. Wussen-Colle M-C, Lalieu C, Simoens C, DeGraef J. Effect of vagotomy and atropine on plasma somatostatin response to a meal in conscious dogs. Regul Peptides. 1988;21:29-36. 23. Melone J. Vagal receptors sensitive to lipids in small intestine of the cat. J Autonom Nerv Syst. 1986;17:231-41. 24. Jeanningros R. Vagal unitary response to intestinal amino acid infusions in the anesthetized cat: a putative signal for proteininduced satiety. Physiol Behav. 1982;28:9-21.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 07:14 For personal use only. No other uses without permission. . All rights reserved.
