Effect of Antisomatostatin y-globulin on Gastrin Release in Rats

Transcription

1 GASTROENTEROLOGY 1981;81:321-6 Effect of Antisomatostatin y-globulin on Gastrin Release in Rats TSUTOMU CHIBA, SEIZO KADOW AKI, TOMOHIKO TAMINATO, KAZUO CHIHARA, YUTAKA SEINO, SHIGERU MATSUKURA, and TAKUO FUJITA The Third Division, Department of Medicine, Kobe University School of Medicine, Kobe, and The Second Division, Department of Medicine, Kyoto University School of Medicine, Kyoto, Japan In order to determine the mechanism of endogenous gastric somatostatin in the regulation of gastrin release, antisomatostatin rabbit y-globulin was administered in vivo and in vitro in the rat. First, in rats anesthetized by chloral hydrate, intravenous injection of 1 ml antisomatostatin rabbit y-globulin had no significant effect on plasma gastrin levels. Second, basal gastrin release from the isolated perfused rat stomach was not affected by the administration of antisomatostatin rabbit y-globulin (1: 99 and 1: 1 dilutions). Third, the addition of antisomatostatin rabbit y-globulin (1: 99) to incubated rat antral mucosa, however, significantly increased the basal gastrin release. From these results, it is concluded that antral somatostatin exerts its inhibitory effect on basal gastrin release mainly not through the circulation but via paracrine pathways. Significant amounts of somatostatin, a potent inhibitor of gastrin release as well as gastric acid secretion, has been found in the fundic and antral mucosa of the stomach by radioimmunoassay as well as by immunocytochemical techniques (1-4). In addition, we have demonstrated the stimulation of somatostatin release by various gastrointestinal peptides from the isolated perfused rat stomach (5-7). Furthermore, it has been recently reported that somatostatin-producing D cells in the antrum have long cytoplasmic processes that terminate on gastrin-producing G cells in rat and humans (8). Although this finding suggests that endogenous gastric somatosta- Received August 15, Accepted March 6, Address requests for reprints to: Tsutomu Chiba, M.D., Third Division of Internal Medicine, Kobe University School of Medicine, Kusunoki-cho, Chuo-ku, Kobe 650, Japan. This work was supported in part by grants from the Ministry of Education, Science and Culture, and from the Ministry of Health and Welfare, Japan by the American Gastroenterological Association /81/ $02.50 tin may playa local regulatory role on gastrin release, the precise mechanism in the regulation of gastrin secretion by endogenous somatostatin is yet uncertain. In an attempt to solve this problem, the effects of antisomatostatin rabbit y-globulin (anti-ssrg) on basal gastrin levels were investigated in vivo and in vitro in the rat. Methods Antisomatostatin y-globulin The neutralizing antisomatostatin serum RB-1105 was produced in the rabbit according to the method described previously (9). At a final dilution of 1: 50,000, this antiserum binds 48% of 125I-N-tyrosyl somatostatin and can detect 12 pmol/l of synthetic somatostatin. The affinity constant (Ka) of this antiserum was 3.91 X lob Llmol, and each milliliter would be capable of binding 5.19 X 10-9 mol somatostatin as calculated on the Scatchard plot (10). The antiserum was specific for somatostatin and showed no cross-reactions with insulin, glucagon, human gastrin I, secretin, rat GH, bombesin, endorphins, vasoactive intestinal polypeptide, substance P, or bovine pancreatic polypeptide. Antisomatostatin serum RB-1105 contained 75 pm of immunoreactive gastrin, while 27 pm of immunoreactive gastrin was found in normal rabbit serum. Thus, in order to avoid the influence of serum constituents other than antisomatostatin antibodies, the y-globulin fraction of the antisomatostatin serum (anti-ssrg) was extracted by two-step precipitation with ammonium sulfate as described elsewhere (11). The y-globulin fraction from normal rabbit serum (NRG) was Similarly obtained. Immediately before use, the Iyophylized y-globulin was dissolved in saline to reconstitute the original concentration in the serum, and further diluted to suitable concentrations with buffer solution in the in vitro experiments. From the binding capacity of this antiserum as described above, 1 ml of anti-ssrg solution diluted to 1: 99 and 1: 1 used in the

2 322 CHIBA ET AL. GASTROENTEROLOGY Vol. B1, No. 2 present in vitro studies would be theoretically able to bind 51.9 pmol and 2.65 nmol somatostatin, respectively, both of which far exceeded the somatostatin levels secreted in the perfusion or in the incubation study. Actually, at dilutions of 1:99 and 1:1 in 1 ml of Krebs-Ringer bicarbonate buffer containing 5.5 mm glucose (KRBG), the anti-ssrg was found to bind 1 pmol synthetic somatostatin within 1 min and within 20 s, respectively, at 37 C (Figure 1). Animals Male Wistar rats, weighing g were used throughout the experiments. The animals were kept in a temperature-controlled and air-conditioned room under artificial lighting (lights on 6:00 AM-6:00 PM), fed Oriental Laboratory Chow (Oriental Yeast Co., Tokyo, Japan), and given tap water ad libitum. In Vivo Study After 24 h fasting, the animals were anesthetized with chloral hydrate (300 mg/kg, Lp.). Thirty minutes later, 1 ml of either NRG or anti-ssrg was injected into the penile vein. Blood sampling (0.8 ml each) was carried out from the jugular vein by the transmuscular puncture technique with a heparinized syringe (12) 5 min, immediately before, and 5, 15, 30, and 45 min after the injection of NRG or anti-ssrg. The blood was put into chilled tubes containing 0.08 ml (800 KIU) Trasylol (Bayer Leverkusen, Germany), and the plasma was immediately separated by centrifugation and kept at -20 C until assayed for gastrin and GH. A small portion of the plasma at each point was used for the detection of the titer of anti-ssrg circulating in the blood. The antibody titer in the rat plasma was determined by measuring its binding capacity for somatostatin on Scatchard plot, as described above. By this procedure, it was found that the rat plasma at each point has 0% (-5 min), 0% (0 min), 16% (5 min), 24% (15 min), 21% (30 min), and 19% (45 min) of the binding capacity of the original antisomatostatin serum (5.19 x 10-9 mol/ml), all of which were sufficiently enought to bind all the circulating soma- tostatin. Indeed, at the dilution of 1: 99 in 1 ml of hormonefree rat plasma, anti-ssrg circulating in the rat blood in any point of time was capable of binding 1 pmol synthetic somatostatin within 3 min at 37 C. Perfusion Study The perfusion of the rat stomach was performed by the method described by Lefebvre et al. (13) as adapted to use in the rat (9). The animals were anesthetized by intraperitoneal injection of 40 mg/kg of pentobarbital after an overnight fast. After a midline laparotomy, polyethylene cannulae were inserted into the left gastric artery and gastric vein. All other vessels and the pancreas were carefully excluded by ligation, and the spleen was then removed. After the esophagocardial junction was ligated, the esophagus was removed, then a catheter was inserted into the stomach through the pyloric ring to drain gastric juice, and finally the duodenum was removed. All perfusions were accomplished with 4.6% dextran (mean mol wt = 70,000) Krebs-Ringer bicarbonate buffer containing 5.5 mm glucose. The perfusate was continuously gassed with 95% O 2-5% CO 2 and maintained at ph 7.4. Both the perfusate and the stomach preparation were kept at 37 C throughout the experiment. Perfusates were administered into the left gastric artery by means of a peristaltic pump at a flow rate of 2 ml/min without recirculation, and anti SSRG or NRG was introduced via side arm to give appropriate final concentrations. After a 20-min perfusion of the stomach with NRG (final concentrations 1: 99 and 1: 1) in buffer solution, anti SSRG (1: 99 and 1: 1) was perfused at a rate of 2 ml/min over a 15-min period. The control experiments were performed with the same dilutions of NRG. The venous effluents were collected every minute into chilled tubes containing a bacitracin-trasylol mixture (2 x 5-5 mol and 1000 KIU/ml, respectively) (bacitracin from Sigma Chemical Co., St. Louis, Mo.). Each sample was divided into two aliquots, and one of them was immediately treated with polyethylene glycol (PEG) for the measurement of free somatostatin; the remaining aliquot was frozen and stored at -20 C until assayed for gastrin. One perfusion was done with each stomach. 100,---:;:--~_----,,-="'----= t 90..e : ~ ~ 70 o z mga I Z 50 w u ~4 a.. Figure 1. Time-course in percent binding of 1 pmol somatostatin to anti-ssrg. One picomole synthetic somatostatin was incubated with 1:199 ~, 1: 99 (0---0) or 1: 1 (e----e) dilution of anti SSRG in 1 ml KRBG at 37 C. There were B tubes in each point of each dilution. All values are the means ± SEM. o sec TIM E 10min

3 August 1981 ANTRAL SOMATOSTATIN AND GASTRIN RELEASE 323 Incubation Study The stomach was quickly removed from rats killed by cervical dislocation after a 24-h fast. The stomach was washed free of food and debris, and the forestomach and corpus were discarded. The mucosa of the antrum was scraped free from the connective tissue and muscle layers by means of glass slides, then immersed in ice-cold KRBG gassed with 95% O 2-5% CO 2 The mucosal scrapings were then placed on filter paper and minced with fine scissors. They were then put into silicon tubes containing 1 ml KRBG gassed with 95% O 2-5% CO 2, and preincubated for 20 min at 37 C. In the first experiment, after the preincubation the mucosal scrapings were transferred to 1 ml KRBG containing NRG (1: 99) and incubated for 30 min (first incubation). The medium was then removed, and 1 ml fresh KRBG containing anti-ssrg or NRG (1: 99) was added, and the mucosal scrapings were incubated for another 30 min (second incubation). In the second experiment, after the first incubation with KRBG alone, 1 ml of fresh KRBG containing various concentrations of synthetic somatostatin (10 2, 10 3, 10', and 105 pm) was added, and the mucosal scrapings were incubated for another 30 min (second incubation). Somatostatin was dissolved in 0.1 mol acetic acid and then diluted with the incubation medium immediately before use. The final ph of the incubation medium was found not to be influenced by the acidic somatostatin solution or vehicle. After the incubation, each medium was immediately removed and put into chilled tubes containing a bacitracin-trasylol mixture (2 X 10-5 mol and 1000 KIU /ml, respectively), frozen, and stored at -20 C until assayed for gastrin. In the first experiment, each medium was divided into two aliquots, and one of them was immediately treated with polyethylene glycol (PEG) for the measurement of free somatostatin. At 4 C in 1 ml KRBG for 30 min, somatostatin and gastrin release from the mucosal scraping were both undetectable, indicating that somatostatin and gastrin release in the incubation study were not due to the damage of mucosal cells. Moreover, at the end of the incubation, histologic examination demonstrated that antral surface cells and mucosal glands were well preserved, although only a moderate edema was observed in a few samples. Endocrine cells were evident with the characteristic morphologic features containing many secretory granules. Hormone Assays Immunoreactive gastrin was measured by the dextran-coated charcoal radioimmunoassay reported previously (14). Antisera raised in a rabbit against synthetic human gastrin I were used for this assay. Synthetic human gastrin I was used as the standard and also as the labeled tracer, and was prepared by a modification of the chloramin-t method (15). The minimal detectable quantity of this assay was 9 pm. The coefficients of variation within and between assays were 6.5% and 8.8%, respectively. Somatostatin did not cross-react with antigastrin antisera. Plasma growth hormone (GH) was measured by a specific radioimmunoassay by using the talcum adsorption technique for separating free and bound hormones, described previously (16,17). Neither NRG nor anti-ssrg interfered with the dextran-coated charcoal RIA method for gastrin, or the talcum adsorption RIA method for rat GH. Measurement of Free Somatostatin Free somatostatin was determined by the method used for free insulin assay (18), with minor modifications. For the measurement of free somatostatin, a 0.03-ml 1.5% y-globulin (Nakarai Chemicals Ltd., Kyoto, Japan) and 0.3-ml 25% (wt/wt) aqueous PEG (Nakarai Chemicals Ltd., Kyoto, Japan) solution (chilled at 4 C) were added to the 0.27-ml sample immediately after collecting the venous effluent or the incubation mixture, mixed vigorously, and centrifuged at 4 C. The supernatant was then assayed for somatostatin. Immunoreactive somatostatin levels were measured by the specific radioimmunoassay by using antiserum RA- 823, described previously (9). Antiserum RA-823 was found not to cross-react with human gastrin I or any other hormones (9). The minimal detectable quantity of the assay was 6 pm. Intra- and interassay variations were 5.4% and 8.5%, respectively. In order to simulate the PEG concentration in the samples, 0.1 mi12.5% PEG was added to the standard tubes. To confirm whether complete precipitation of the somatostatin antibodies was accomplished with PEG, the supernatants were checked as follows: 0.1 ml 125I-N-tyrosyl somatostatin was added to 0.4 ml buffer containing 0.1 ml of the PEG supernatant. After a two-day incubation at 4 C, free 125I-N-tyrosyl somatostatin was adsorbed with dextran-coated charcoal. By this procedure it was found that complete precipitation of the antibodies was accomplished in all samples. Analysis of the Data In the incubation study, the amounts of gastrin and free somatostatin detected in the second incubation medium were expressed as percentages of the amounts detected in the first incubation medium. All results were expressed as mean ± SEM. Statistical analysis was performed by the unpaired Student's t-test when two means were compared, and by Duncan's multiple-range test when more than two means were compared (19). For comparison within the group, the paired t test was used. Origin of Materials The N-tyrosyl somatostatin was kindly given by Dr. D. H. Coy (V.A. Hospital, New Orleans, La). The human gastrin I was donated by I.C.I., Ltd. (Macclesfield Cheshire, United Kingdom). Synthetic somatostatin was purchased from Protein Research Institute (Osaka, Japan). Antiserum 30K was purchased from the Diabetes Fund

4 324 CHIBA ET AL. GASTROENTEROLOGY Vol. 81. No.2 (University of Texas Southwestern Medical School, Dallas, Tex.). Results 60 "0 ~4 z I In Vivo Study Under chloral hydrate anesthesia, the mean basal levels of plasma gastrin and GH before the injection of anti-ssrg or NRG were 35 ± 5 pm and 0.47 ± 0.10nM, respectively. Plasma gastrin levels increased gradually but not significantly after the intravenous injection of 1 ml anti-ssrg or NRG. In addition, no significant difference in plasma gastrin levels was found between anti-ssrg-injected rats and NRG-injected rats, as shown in Figure 2. In contrast, intravenous injection of anti-ssrg evoked an 10-fold increase in plasma GH levels 15 min after the injection, while plasma GH was not changed significantly by the administration of NRG. The peak value of plasma GH after anti-ssrg injection was 4.61 ± 0.89 nm, significantly higher than the corresponding values in NRG-injected rats (0.56 ± 0.20 nm, p < 0.01) (Figure 2). o 5 _ 4 o E 3 c: I 2 <.!) - --1'"" <r 'f' "? 0_ T I ME(mln) Figure 2. Effects of anti-ssrg on plasma gastrin and GH levels in the rat. After chloral hydrate anesthesia, 1 ml either anti-ssrg (e e) or NRG (0---0) was injected into the penile vein. There were 6 animals in each group. Each value represents the mean ± SEM. 'Significantly different (p < 0.01) from NRG. Perfusion Study The mean basal free somatostatin levels in the venous effluents before the stimulation period were 110 ± 8 pm and 106 ± 7 pm (± SEM), and gastrin levels were 82 ± 7 pm and 84 ± 5 pm (± SEM) when the perfusions were performed with buffer solution containing NRG (1: 99 and 1: 1 dilutions, respectively), neither of which changed significantly throughout the perfusion period in the control experiment. On the other hand, free somatostatin disappeared after the administration of anti-ssrg (1: 99 and 1: 1). However, anti-ssrg (1: 99 and 1: 1) caused no significant change in gastrin secretion. The total amounts of gastrin and free somatostatin detected during the 15-min perfusion with anti SSRG or NRG are summarized in Table 1. Incubation Study In the first experiment, gastrin and free somatostatin levels in the first incubation medium were 275 ± 45 pm and 82 ± 13 pm, respectively, both of which were unchanged through the first and second incubation with KRBG containing NRG (1: 99). In contrast, free somatostatin became undetectable when incubated with KRBG containing anti-ssrg (1: 99), while gastrin levels were increased significantly, compared with the values when incubated with the same concentration of NRG (p < 0.01) (Table 2). In the second experiment, a and 105-pM somatostatin administration decreased gastrin release significantly (72 ± 9%, P < 0.05 and 58 ± 12%, P < 0.01, respectively) compared with the controls (94 ± 5%), while a and 103-pM somatostatin administration had no significant effect on gastrin release (92 ± 15% and 91 ± 14%, respectively). Discussion Somatostatin is well known to inhibit gastrin release (20,21), and Larsson et al. (8) have demonstrated recently that D cells in the antrum have long cytoplasmic processes that are in direct contact with the gastrin-producing G cells. Although this suggests a direct paracrine role for gastric somatostatin in the regulation of gastrin release, clear evidence of this paracrine action has not been reported. In the present study we neutralized the effects of endogenous somatostatin by the administration of a specific antisomatostatin y-globulin (anti-ssrg) in vivo and in vitro in rats so as to elucidate this regulatory mechanism. The present study shows that Lv. injection of anti SSRG has no significant effect on basal plasma gastrin levels. In addition, the administration of anti SSRG causes no significant change in basal gastrin release from the isolated perfused rat stomach. The anti-ssrg in the present experiment was obviously effective in neutralizing the circulating somatostatin

5 August 1981 ANTRAL SOMATOSTATIN AND GASTRIN RELEASE 325 Table 1. Total Amounts of Free Somatostatin and Gastrin Released During the Infusion of Antisomatostatin Rabbit y-globulin or Normal Rabbit y-globulin from Isolated Perfused Rat Stomach NRG 1: 99 Dilution anti-ssrg 1: 1 Dilution Free somatostatin (pmo1!15 min) Gastrin 3.55 ± ± 0.36 NO (pmol/15 min) 2.48 ± ± ± ± 0.34 NRG anti-ssrg There were 6 animals for each group. All values represent the means ± SEM of the total amounts of free somatostatin or gastrin released during the infusion of NRG or anti-ssrg (1 : 99 and 1: 1 dilutions) for 15 min. a NO = not detected. because the titer of anti-ssrg circulating in the blood of rats at any point seemed to be enough to bind all the circulating somatostatin and the GH level increased significantly after anti-ssrg administration in the study in vivo, as have been previously reported (22,23), and free somatostatin was found to be undetectable during the administration of the anti-ssrg in the perfusion experiment. As injected antibody tends not to reach the interstitial spaces easily to inactivate local somatostatin (24), the ineffectiveness of anti-ssrg on gastrin release both in vivo and in the perfusion study strongly suggests that circulating somatostatin does not suppress gastrin release via either endocrine pathways or local circulation under physiologic conditions. In contrast, the addition of anti-ssrg to incubated rat antral mucosa evoked a significant increase in gastrin release. As somatostatin might accumulate to a greater extent in the in vitro incubations than would occur in the stomach under physiologic conditions, we considered whether the neutralizing effect of the anti-ssrg as measured in the first experiment of the incubation study would seem to be exaggerated. We found, however, that, in the second experiment of the incubation study, 100 pm of somatostatin, almost the same level released into the incubation medium in the first experiment, did not have any inhibitory effect on gastrin release, although endogenous gastric somatostatin in the rat may not have equally bi- ologic activity as synthetic somatostatin. It is possible, therefore, that antral somatostatin exerts a tonic inhibitory effect on basal gastrin secretion via paracrine pathways. Recently, Schusdziarra et al. (24) reported that neutralization of circulating somatostatin increases plasma concentrations of gastrin in dogs. Saffouri et al. (25) have reported that infusion of antisomatostatin serum produces significant increases in gastrin secretion from the isolated perfused rat stomach. The reason for these different results is not known. As to the in vivo study, the difference in species or the characteristics of the antisomatostatin serum used might be important. In the perfusion study, the difference in the characteristics of the antiserum may also be responsible for the different results. Furthermore, as our antisomatostatin serum (RB-ll05) contained not a little amount of gastrin, we used the extracted antisomatostatin y-globulin instead of crude antiserum to avoid the influence of serum constituents other than antisomatostatin antibodies. In any case, the present study lends support to the histologic data of Larsson et al. (8), and emphasizes the physiologic significance as a paracrine hormone of antral somatostatin on gastrin release. References Table 2. Free somatostatin ( %) Gastrin ( %) Effect of Antisomatostatin Rabbit y-globulin (1; 99) on Gastrin and Free Somatostatin Release From Incubated Rat Antral Mucosa NRG 94 ± ± 8 anti-ssrg Number of tubes were 8 for each group. The amounts of gastrin and free somatostatin detected in the second incubation medium were expressed as percentage of the amounts detected in the first incubation medium. All values represent the means ± SEM. a NO = not detected. b Significantly different (p < 0.01) from percent gastrin in NRG. 1. Arimura A, Sato H, Dupont A, et al. Somatostatin: abundance of immunoreactive hormone in rat stomach and pancreas. Science 1975;189: Patel YC, Reichlin S. Somatostatin in hypothalamus. extrahypothalamic brain and peripheral tissues of the rat. Endocrinology 1978;102: Polak JM. Pearse AGE, Grimelius L. et al. Growth-hormone release inhibiting hormone in gastrointestinal and pancreatic o cell. Lancet 1975;1: Dubois PM. Paulin C. Dubois MP. Gastrointestinal somatostatin cells in the human fetus. Cell Tiss Res 1976;166: Chiba T. Taminato T. Kadowaki S. et al. Effects of various gastrointestinal peptides on gastric somatostatin release. Endocrinology 1980;106: Chiba T, Taminato T. Kadowaki S. et al. Effects of [Asu ' 7]-eel

6 326 CHIBA ET AL. GASTROENTEROLOGY Vol. 81. No.2 calcitonin on gastric somatostatin and gastrin release. Gut 1980;21: Chiba T. Taminato T. Kadowaki S. et al. Effects of glucagon. secretin. and vasoactive intestinal polypeptide on gastric somatostatin and gastrin release from isolated perfused rat stomach. Gastroenterology 1980;79: Larsson LI. Goltermann N. Magistris L. et a1. Somatostatin cell processes as pathways for paracrine secretion. Science 1979;205: Chiba T. Seino Y. Goto Y. et al. Somatostatin release from isolated perfused rat stomach. Biochem Biophys Res Commun 1978;82: Scatchard G. The attractions of proteins for small molecules and ions. Ann NY Acad Sci 1949;51: Weir DM. Handbook of experimental biology. Philadelphia: FA Davis. 1967; Kato Y. Dupre J, Beck J. Plasma growth hormone in the anesthetized rat. Endocrinology 1973;93: Lefebvre PJ. Luyckx AS. Factors controlling gastric-glucagon release. J Clin Invest 1977;59: Seino Y. Matsukura S. Miyamoto Y. et al. Hypergastrinemia in hyperthyroidism. J Clin Endocrinol Metab 1976;43: Hunter WM. Greenwood FC. Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature (Lond) 1962;194: Kato Y. Chihara K. Maeda K. et al. Plasma growth hormone responses to thyrotropin-releasing hormone in the urethaneanesthetized rat. Endocrinology 1975;96: Chihara K. Kato Y. Ohgo S. et al. Stimulating and inhibiting effects of thyrotropin-releasing hormone on growth hormone release in rats. Endocrinology 1976;98: Kuzuya H. Blix PM. Horwitz DL. et al. Determination of free and total insulin and C-peptide in insulin-treated diabetics. Diabetes 1977;26: Steel RDG. Torrie JH. In: Principles and procedures of statistics. New York: McGraw-Hill. 1966: Bloom SR. Mortimer CH. Thorner MO. et a1. Inhibition of gastrin and gastric acid secretion by growth hormone release inhibiting hormone 1974;Lancet 2: Hayes JR. Johnson DG. Koerker D. et al. Inhibition of gastrin release by somatostatin in vitro. Endocrinology 1975;96: Terry LC. Willoughby JO. Patel Y. Antiserum to somatostatin prevents stress-induced inhibition of growth hormone secretion in the rat. Science 1976;192: Chihara K. Arimura A. Chihara M. et al. Studies on the mechanism of growth hormone and thyrotropin responses to somatostatin antiserum in anesthetized rats. Endocrinology 1978;103: Schusdziarra V. Zyznar E. Rouiller D. et al. Splanchnic somatostatin: a hormonal regulator of nutrient homeostasis. Science 1980;207: Saffouri B. Weir G. Bitar K. et al. Stimulation of gastrin secretion from the perfused rat stomach by somatostatin antiserum. Life Sci 1979;25: