Effect of Sepsis or Cytokine Administration Release of Gut Peptides Oded Zamir, MD, Per-Olof Hasselgren, MD, PhD, FAG, Takashi Higashiguchi, Janice A. Frederick, MD, Josef E. Fischer, MD, FACS, Cincinnati,Ohio
The effect of sepsis on plasma levels of various gut peptides was studied in rats. Sepsis was induced by cecal ligation and puncture (CLP) ; control animals underwent sham operation. Sixteen hours after CLP or sham operation, portal and systemic blood was drawn, and plasma levels of gastrin, vasoactive intestinal peptide (VIP), secretin, peptide YY (PYY) , gastrin-releasing peptide (CRP) , and suhstance P were determined by radioimmunoassay. Plasma levels of gastrin, VIP, PYY, and secretin were elevated in septic rats compared with nonseptic animals, with the highest levels noted in portal blood. There was no effect of sepsis on GRP or substance P levels. In other experiments, human recombinant interleulcin la (IL-la) or recombinant tumor necrosis factor Q! (TNFa) was injected intraperitoneally (300 cLg/lcgbody weight in 3 divided doses over 16 hours). There was no change in plasma levels of gut peptides after IL-la injection. TNFo induced elevation of PYY levels in portal plasma with no change in other gut peptide levels. The results suggest that sepsis stimulates release of certain gut peptides and that TNF, but not IL-l, may be partly responsible for this response. The mechanism of the release of gut peptides and its significance in the pathophysiologic changes induced by sepsis remain to he determined.
MD,
epsis induces profound hemodynamic and metabolic responses in various organs and tissues. It is only S recently that the gut has been recognized as playing an important role in this response, In a recent study from our laboratory, intestinal protein synthesis in rats was increased during sepsis [I], whereas at the same time, the protein content of the intestinal wall was unchanged. One possible explanation for the unchanged protein content of the intestinal wall, despite the increased protein synthesis, may be increased synthesis of secreted proteins. Previous studies showed that plasma levels of vasoactive intestinal peptide (VIP) were elevated during sepsis or endotoxemia [2-d], results that may be consistent with increased intestinal release of the peptide. It is not known if this response is specific for VIP or if other peptides are affected by sepsis in a similar manner. Therefore, in the current study, we examined the influence of sepsis on plasma levels of a number of peptides that are believed to originate mainly in the gastrointestinal tract. These included gastrin, produced mainly in the antral mucosa [5]; VIP, a peptide that is widely distributed along the gastrointestinal tract [q; secretin, a peptide structurally related to VIP and found mainly in duodenal and jejunal mucosa [7]; peptide YY (PYY), concentrated in the mucosa of the distal ileum and colon [8]; and gastrin-releasing peptide (GRP) [9] and substance P [IO], which are also found at high levels in the gut. Because various cytokines, in particular tumor necrosis factor (TNF) and interleukin-l (IL-l), have been implicated as mediators of different metabolic changes during sepsis [II], we also measured plasma levels of gut peptides after treatment of rats with TNF or IL-l. MATERIALS
From the Department of Surgery, University of Cincinnati Medical Center, Cincinnati, Ohio. Requests for reprints should be addressed to Per-Olaf Hasselgren, MD, Department of Surgery, University of Cincinnati, 231 Bethesda Avenue, Ciinnati, Ohio 45267-0558. Presented at the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orleans, Louisiana, May 20-22,199l.
on
AND METHODS
Studies were performed in male Sprague-Dawley rats (Zivic-Miller, Zelienople, PA) weighing 150 to 200 g. Rats were housed two per cage at 22OC in a room with a 12-hour light/dark cycle. Two series of experiments were performed. In the first series of experiments, the effect of sepsis on plasma levels of gut peptides was studied. Sepsis was induced by cecal ligation and puncture (CLP) as described previously 1121. Control animals underwent sham operations. Rats were resuscitated with normal saline (5 mL/ 100 g body weight) administered subcutaneously on the back at the time of operation. To avoid any influence of different food intake on metabolic changes, both groups of rats were fasted but had free access to water following the operative procedures. Sixteen hours after CLP or sham operation, animals were anesthetized with ether and 3 to 4 mL of blood was drawn from the right atrium (referred to as systemic blood) and the portal vein. The blood was collected into chilled heparinized
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
181
TABLE I Portal and Systemic Plasma Concentrations @g/ml) of Gut Peptides in Control (n = 7) and Septic (n = 9) Rats Systemic
VIP PW Secretin Substance P Gastrin GRP
Portal
Sham
CLP
3.0 -+ 1.9 32.1 2 5.0 23.5 lr 4.4 ND ND ND
40.7 + 13.4* 73.4 + a.o* 95.3 f 14.7* ND ND ND
CLP
Sham 16.4 46.3 30.1 66.6 0.96 16.3
2 + + + 2 2
4.6f 4.6 5.2 13.7 0.44 3.2
213.8 ” 46.2*+ 108.4 k i7.2*+ 127.2 100.1 14.6 14.7
2 13.5* * 13.7 f 2.6* + i .a
CLP = cacal ligation and puncture; VIP = vasoactive intestinal peptfde; PYY = peptide w; GRP = gaatrin-releasingpeptide; ND = not determined. *p < 0.05 varsua sham. tp ~0.05 varaus systemic blood in the corresponding group.
TABLE II Portal and Systemic Plasma Concentrations (pglml) of Gut Peptides In Control (n = 6 to 10) and TNF-Treated (n = 7 to 10) Rats Portal Vein
VIP PW Secretin
Systemic Blood
Control
TNF
Control
TNF
42.3 + 4.5 52.6 2 6.7 67.0 + 9.6
53.7 ? 11.7 156.6 f 26.2* 75.7 f 15.0
24.4 -+ 3.5 66.4~ 10.1 69.0 + 6.5
26.6 -t 5.9 99.1 2 10.1 69.5 f 21.2
TNF = tumor necrosis factor; VIP = vasoactiva intestinal peptide; PW = peptide ‘r-f. lp co.05 versus control.
TABLE III Portal Plasma Concentrations @g/ml) of Gut Peptides in Control (n = 7 or 6) and IL-l-Treated Rats (n = 6 to 9) Control VIP PW Secretin Substance P Gastrin GRP
95.1 103.4 41.4 33.6 12.5 16.7
2 5.6 + 19.5 2 6.0 -+ 6.1 * 3.9 + 2.4
IL-1 76.1 134.1 23.5 31.6 21.1 12.4
+ t f + 2 +
5.7 15.2 3.0 4.6 4.1 1.6
IL-1 = Interleukin-1; VIP = vasoactive intestinal peptide; PW = peptide W; GRP = gastrin-releasingpeptide.
glass tubes containing aprotinin (400 KIU/mL blood, Sigma Inc., St. Louis, MO) and centrifuged at 3,500g for 15 minutes at 4%. Plasma was stored at -7O’C until assayed (see below). Animals were studied 16 hours after induction of sepsis in the current report because, in previous studies, changes in protein metabolism, including increased intestinal protein synthesis, were observed at this time point [I ,I 21. In the second series of experiments, the effect of TNF or IL-1 on plasma levels of gut peptides was examined. Human recombinant TNFa (rTNFar) was kindly provided by Knoll Pharmaceuticals, Whippany, NJ, and had a 182
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
specific activity of more than 5 X lo6 U/mg protein (L929 cell cytotoxicity) and an endotoxin concentration of less than 5 EU/mg protein (lymulus lysate assay). The cytokine was dissolved in 5 mM phosphate-buffered saline (PBS), pH 7.4, to a final concentration of 10 pg/mL. Human recombinant IL- la (rIL- 1a) was kindly provided by Hoffmann-La Roche, Inc., Nutley, NJ, and had a specific activity of 3 X lo* U/mg protein (D-10 assay) and an endotoxin concentration leas than 0.75 EU/mg protein (lymulus lysate assay). The cytokine was dissolved in PBS to a final concentration of 10 pg/mL. Recombinant TNFa or rIL- 1a! was injected intraperitoneally iKith a total dose of 300 pg/kg body weight, divided into three equal doses with 8-hour intervals. Control animals received corresponding volumes of PBS. This protocol of TNF or IL-l treatment resulted in metabolic changes in skeletal muscle in previous experiments in our laboratory [ 13,14], Rats had free access to water but food was withheld after the first cytokine or control injection in order to avoid any influence of reduced food intake on metabolic changes caused by the cytokines. Two hours after the last cvtokine or control injection, animals were anesthetized with ether, and blood samples were drawn and handled as described above for septic rats. Plasma levels of VIP, substance P, gastrin, GRP, and secretin were determined by radioimmunoassay (RIA), using kits from Peninsula Laboratories (Belmont, CA). RIA of PYY was performed as described previously [ 151, using synthetic porcine PYY labeled with Na1251by the chloramine-T method and antiserum raised in rabbits immunized against porcine PYY [16]. The minimal level of PYY detected with this technique is 20 pg/mL. Results are expressed as mean f SEM. Student’s ttest or analysis of variance followed by Duncan’s test was used for the statistical analysis. RESULTS In the first series of experiments, the mortality rate 16 hours after CLP was 25% (6 of 24). No control rats died. All rats surviving CLP showed signs of sepsis including lethargy, piloerection, diarrhea, and discharge around the nostrils. Basal plasma concentrations of VIP were higher in portal than in systemic blood (Table I). Sepsis resulted 163
JANUARY
1992
EFFECT OF SEPSIS ON GUT PEPTIDE
in increased plasma levels of VIP, PYY, secretin, and gastrin, with the highest concentrations found in portal blood (Table I). Sepsis did not significantly affect portal plasma concentrations of substance P or GRP, and these peptides were not measured in systemic blood. In the second series of experiments, there was no mortality following either TNF or IL-1 injections, but the cytokinetreated animals showed moderate lethargy, piloerection, and diarrhea. Administration of TNF resulted in increased plasma levels of PYY in portal blood but no other significant changes were noted in VIP, PYY, or secretin levels (Table II). Substance P, gastrin, and GRP were not measured in this experiment. In the experiment in which IL-l was administered, only portal vein plasma concentrations of the gut peptides were measured. IL-l did not significantly affect plasma levels of any of the peptides measured (Table RI). It should be noted that control plasma levels of the gut peptides were generally lower in the first series of experiments (Table I) than in the second series of experiments (Tables II and III), both in portal and systemic blood. This may reflect the fact that control rats underwent different procedures in the different experimental series and that the experiments were performed at different times. It is important to point out, however, that control and experimental groups were always studied at the same time within the different experiments. COMMENTS
The current study showed that sepsis in rats resulted in increased plasma levels of the gut peptides VIP, PYY, secretin, and gastrin. Thus, the response to sepsis does not seem to be specific for VIP, which was already shown in previous reports to be increased during sepsis [4] or endotoxemia [2,3]. On the other hand, sepsis probably does not stimulate the release of all gastrointestinal peptides because, in the current study, plasma levels of substance P and GRP were not affected by sepsis. Although increased plasma levels of gut peptides during sepsis, in particular in portal blood, may be consistent with increased intestinal release into the circulation, the results may also reflect reduced clearance of the peptides, especially in view of their short half-life times of a few minutes only. It should also be noted that a reduction in portal blood flow could give rise to increased portal concentrations of the peptides even if production rates were unchanged. Therefore, measurements of portal blood flow would have been necessary to calculate actual release rates of the peptides. Previous studies from our laboratory demonstrated that rats are hemodynamically stable 16 hours after CLP [Z7]. In other reports, intestinal blood flow in rats was increased during septic peritonitis [18] or unchanged during endotoxemia [19]. Thus, it is unlikely that the current results of increased portal plasma concentrations of the peptides reflected decreased blood flow. For the same reason, it is unlikely that gut ischemia was responsible for increased release of the pep tides as suggested previously for VIP [2U]. Increased release of gut peptides during sepsis can result from increased synthesis of the peptides and/or
stimulated release of preexisting intracellular pools. In recent experiments in our laboratory, intestinal protein synthesis was increased in septic rats with no concomitant change of protein content in the intestinal wall [I], a result that may be consistent with stimulated synthesis of secreted proteins. Thus, increased release of gut peptides during sepsis may at least partly reflect stimulated synthesis of the substances. The current results suggest that the release of gut peptides during sepsis is not primarily mediated by TNF or IL- 1, although TNF may be partly responsible for the release of PYY. It is important to point out, however, that the results do not rule out the possibility that TNF and IL-1 regulate the release of gastrointestinal hormones during sepsis. Cytokines may have different biologic effects in the septic organism than in healthy rats since various factors present during sepsis may act as important co-factors to the cytokines. It is also possible that the amounts of cytokines administered here were not large enough to induce sepsis-like plasma or tissue concentrations of the cytokines. It should be noted, however, that the same amounts of TNF or IL- 1 as used here resulted in metabolic alterations in skeletal muscle that were almost identical to those observed during sepsis [13,14]. Further experiments, utilizing specific receptor blockers or cytokine antibodies, are necessary to test the hypothesis that cytokines regulate the release of gastrointestinal hormones during sepsis. The biologic significance of elevated plasma levels of gut peptides during sepsis is unknown and can only be speculated on at present. Some of the physiologic alterations induced by sepsis may be mediated or affected by gut peptides. For example, gastric erosions or ulcerations during sepsis may be caused by gastrin-induced acid secretion [21]. Gut peptides have been implicated in delayed gastric emptying [22,23] and impaired intestinal motility [24-24 and may play a role in the pathogenesis of motility abnormalities induced by sepsis. Some metabolic changes characteristic of sepsis, such as glycogenolysis and lipolysis, can be elicited by VIP [27,28]. This peptide may also affect the hemodynamic changes induced by sepsis through its vasodilatory action [29] and positive ionotropic [30] effects. Other peptides may have a beneficial effect during sepsis by enhancing intestinal mucosal growth [31,32]. In summary, the current results suggest that sepsis induces increased plasma levels of most, but not all, gastrointestinal hormones. Except for PYY, a similar response was not found after TNF or IL-l administration. Further studies are necessary to define the mechanisms and biologic implications of gut peptide release in sepsis. REFERENCES 1. Von Allmen D, Hasselgren PO, Higashiguchi T, Zamir 0, Fischer JE. Increased intestinal protein synthesis during sepsis. Surg Forum 1990; 41: 68-70. 2. Freund H, Ebeid AM, Fischer JE. An increase in vasoactive intestinal peptide levels in canine endotoxin shock. Surg Gynecol Obstet 1981; 152: 604-6. 3. Revhaugh A, Lygren I, Lundgren TI, Andersen OK, Burhol PG,
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
183
23. Valenzuela JE, Defilippi C. Inhibition of gastric emptying in humans by secretin, the cctapeptide of cholecystokinin, and intraduodenal fat. Gastroenterology 1981; 81: 898-902. 24. Mukhopadhyay AK, Johnson LR, Copeland EM, Weisbrodt NW. Effect of secretin on electrical activity of the small intestine. Am J Physiol 1975; 229: 484-8. 25. Kachelhoffer J, Mendel C, Dauchel J, Hohmatter D, Grenier JF. The effect of VIP on intestinal motility. Study on ex uiuo perfused isolated canine jejunal loops. Am J Dig Dis 1976; 21: 957-62. 26. Krantis A, Potwin W, Harding RK. Peptide YY (PYY) stimulates intrinsic enteric motor neurons in the rat small intestine. Arch Pharm (Weinheim) 1988; 338: 287-92. 27. Kevins C, Said SI. Hyperglycemic and glycogenolytic effects of vasoactive intestinal polypeptide. Proc Sot Exp Biol Med 1973; 142: 1044-7. 28. Frandsen EK, Moody AJ. Lipolytic action of newly isolated vasoactive intestinal polypeptide. Horm Metab Res 1973; 5: 196-9. 29. Said SI, Mutt V. Potent peripheral and splanchnic vasodilator peptide from normal gut. Nature 1970; 225: 863-4. 30. Said SI, Mutt V. Polypeptide with broad biological activity: isolation from small intestine. Science 1970; 169: 1217-8. 31. Bloom SR. Gut hormones in adaptation. Gut 1987; 28: 31-5. 32. Johnson LR, Lichtenberger LM, Copeland EM, Dudrick SJ, Castro GA. Action of gastrin on gastrointestinal structure and function. Gastrcenterology 1975; 68: 1184-92.
Giercksky KE. Increased plasma levels of vasoactive intestinal polypeptide in pigs during endotoxemia. Em Surg Res 1985; 17: 75-82. 4. Fuortes M, Blank MA, Scalea TM, Pollock TW, Jaffe BM. Release of vasoactive intestinal peptide during hyperdynamic sepsis in dogs. Surgery 1988; 104: 894-8. 5. Weiner I, Khalil T, Thompson JC, Rayford PL. Gastrin. In: Thompson JC, Greely GH, Rayford PL, Townsend CM, editors. Gastrointestinal endocrinology. New York: McGraw Hill, 1987; 194-212. 6. Besson J, Laburthe M, Bataille D, Dupont C, Rosselin G. Vasoactive intestinal peptide (VIP): tissue distribution in the rat as measured by radioimmunoassay and by radioreceptorassay. Acta Endccrinol 1978; 87: 799-810. 7. Straus E, Yalow RS. Immunoreactive secretin in gastrointestinal muccsa of several mammalian species. Gastroenterology 1978; 75: 401-4. 8. Lundberg SM, Tatemoto K, Terenius L, et al. Localization of peptide YY (PYY) in gastrointestinal endocrine cells and effects on intestinal blood flow and motility. Proc Nat1 Acad Sci U S A 1982; 79: 4471-5. 9. Greeley GH, Spannagel A, Burdet JB. Distribution of gastrinreleasing peptide and bombesin-like peptides in the alimentary canal of rats, rabbits, dogs and humans [abstract]. Gastroenterology 1984; 86: 1097. 10. Holzer P, Bucsics A, Saria A, Lembeck F. A study of the concentration of substance P and neurotensin in the gastrointestinal tract of various mammals. Neuroscience 1982; 7: 2919-24. 11. Flores EA, Bistrian BR, Pomposelli JJ, et al. Infusion of tumor necrosis factor/cache&in promotes muscle catabolism in the rat. A synergistic effect with interleukin-1. J Clin Invest 1989; 83: 1614-22. 12. Hasselgren PO, Talamini MA, James JH, et al. Protein metab olism in different types of skeletal muscle during early and late sepsis in rata. Arch Surg 1986; 121: 918-23. 13. Zamir 0, Hasselgren PO, von Allmen D, Fischer JE. The effect of interleukin-1 and the glucocorticoid receptor blocker RU 38486 on total and myotibrillar protein breakdown in skeletal muscle. J Surg Res 1991; 50: 579-83. 14. Zamir 0, Hasselgren PO, Kunkel SL, Higashiguchi T, Frederick JA, Fischer JE. Evidence that TNF participates in the regulation of muscle proteolysis during sepsis. Arch Surg. In press. 15. Chen MH, Balasubramaniam A, Murphy RF, et al. Sensitive radioimmunoassay for measurement of circulating peptide YY. Gastroenterology 1984; 87: 1332-8. 16. Balasubramaniam A, O’Hare MMT, Murphy RF, Joffe SN, Buchanan KD, Fischer JE. Synthesis of PYY for production of antibodies [abstract]. Regul Pept 1984; 7: 275. 17. Pedersen PV, Warner BW, Bjornson HS, et al. Hemodynamic and metabolic alterations during experimental sepsis in young and adult rats. Surg Gynecol Obstet 1989; 168: 148-56. 18. Lang CH, Bagby GJ, Ferguson JL, Spitzer JJ. Cardiac output and redistribution of organ blood flow in hypermetabolic sepsis. Am J Physiol 1984; 246: R331-7. 19. Jepson MM, Cox M, Bates PC, et al. Regional blood flow and skeletal muscle energy status in endotoxemic rata. Am J Physiol 1987; 252: E581-7. 20. Modlin IM, Bloom SR, Mitchell S. Plasma vasoactive polypeptide (VIP) levels and intestinal ischaemia. Experientia 1978; 34: 535-6. 21. Gdonkor P, Mowat C, Himal HS. Prevention of sepsis induced gastric lesions in dogs by cimetidine via inhibition of gastric secretion and by prostaglandin via cytoprotection. Gastroenterology 1981; 80: 375-9. 22. Debas HT, Farooq D, Grossman MI. Inhibition of gastric emptying is a physiological action of cholecystokinin. Gastroenterology 1975; 68: 1211-7.
When you give these cytokines intraperitoneally, how are they absorbed? Are they absorbed quickly or over a period of several hours? How is intraperitoneal administration compared with intravenous injection? Haile T. Dehas (San Francisco, CA): I noticed that the levels of the two neuropeptides, substance P and gastrin-releasing peptide, were not raised. Is it possible that there was actually no increase in release, but that there was a reduction in mesenteric blood flow? Did you measure portal blood flow? William Silen (Boston, MA): My question has to do with whether you determined if this was simply a release of stored gastrin rather than stimulation of formation. In other words, was there destruction or injury of the intestinal mucosa or other portions of the gastrointestinal tract that might have stored peptides that would be released into the portal circulation during shock? Mark Evers (Galveston, TX): Have you measured actual peptide YY content? Are you increasing new peptide synthesis or simply stimulating release in response to sepsis? Oded Zamir (closing): Selection of the doses of recombinant interleukin- 1 (rIL- 1(Y)and recombinant tumor necrosis factor (rTNF-ol) was based on previous studies in our laboratory, which showed changes in protein metabolism in skeletal muscle and gut after administration of the cytokines in the same dose. It should be noted, however, that it is unknown if the circulating, and more importantly, tissue levels of the cytokines following injection at this dose are similar to those seen in sepsis or endotoxemia. Absorption of the cytokines following intraperitoneal injection, as reflected by the metabolic and endocrine
184
163
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
DISCUSSION Dr. Cheung:
JANUARY
1992
EFFECT OF SEPSIS ON GUT PEPTIDE
response, is very rapid. For example, we found a striking elevation of plasma corticosterone as early as 30 minutes after intraperitoneal injection of either rIL-la or rTNF-a. Dr. Debas, we have not measured portal or mesenteric blood flow in the current study. However, other investigators found mesenteric blood flow to be unchanged or increased after peritonitis or endotoxemia in rats. Thus, elevated portal levels probably reflect increased release of the peptides rather than higher concentrations due to reduced blood flow after sepsis.
THE AMERICAN
Dr. Silen, this is an important question but, from the current data, we cannot exclude the possibility that elevated plasma levels of the peptides reflect release of stored peptides rather than increased synthesis. Dr. Evers, we did not measure tissue content of gut peptides. However, the fact that plasma levels of these very short-lived peptides were elevated 16 hours after induction of sepsis supports the idea that elevated plasma levels reflect increased synthesis in the gut; further studies are necessary to address this question.
JOURNAL
OF SURGERY
VOLUME
163
JANUARY
1992
185
The effect of sepsis on plasma levels of various gut peptides was studied in rats. Sepsis was induced by cecal ligation and puncture (CLP) ; control animals underwent sham operation. Sixteen hours after CLP or sham operation, portal and systemic blood was drawn, and plasma levels of gastrin, vasoactive intestinal peptide (VIP), secretin, peptide YY (PYY) , gastrin-releasing peptide (CRP) , and suhstance P were determined by radioimmunoassay. Plasma levels of gastrin, VIP, PYY, and secretin were elevated in septic rats compared with nonseptic animals, with the highest levels noted in portal blood. There was no effect of sepsis on GRP or substance P levels. In other experiments, human recombinant interleulcin la (IL-la) or recombinant tumor necrosis factor Q! (TNFa) was injected intraperitoneally (300 cLg/lcgbody weight in 3 divided doses over 16 hours). There was no change in plasma levels of gut peptides after IL-la injection. TNFo induced elevation of PYY levels in portal plasma with no change in other gut peptide levels. The results suggest that sepsis stimulates release of certain gut peptides and that TNF, but not IL-l, may be partly responsible for this response. The mechanism of the release of gut peptides and its significance in the pathophysiologic changes induced by sepsis remain to he determined.
MD,
epsis induces profound hemodynamic and metabolic responses in various organs and tissues. It is only S recently that the gut has been recognized as playing an important role in this response, In a recent study from our laboratory, intestinal protein synthesis in rats was increased during sepsis [I], whereas at the same time, the protein content of the intestinal wall was unchanged. One possible explanation for the unchanged protein content of the intestinal wall, despite the increased protein synthesis, may be increased synthesis of secreted proteins. Previous studies showed that plasma levels of vasoactive intestinal peptide (VIP) were elevated during sepsis or endotoxemia [2-d], results that may be consistent with increased intestinal release of the peptide. It is not known if this response is specific for VIP or if other peptides are affected by sepsis in a similar manner. Therefore, in the current study, we examined the influence of sepsis on plasma levels of a number of peptides that are believed to originate mainly in the gastrointestinal tract. These included gastrin, produced mainly in the antral mucosa [5]; VIP, a peptide that is widely distributed along the gastrointestinal tract [q; secretin, a peptide structurally related to VIP and found mainly in duodenal and jejunal mucosa [7]; peptide YY (PYY), concentrated in the mucosa of the distal ileum and colon [8]; and gastrin-releasing peptide (GRP) [9] and substance P [IO], which are also found at high levels in the gut. Because various cytokines, in particular tumor necrosis factor (TNF) and interleukin-l (IL-l), have been implicated as mediators of different metabolic changes during sepsis [II], we also measured plasma levels of gut peptides after treatment of rats with TNF or IL-l. MATERIALS
From the Department of Surgery, University of Cincinnati Medical Center, Cincinnati, Ohio. Requests for reprints should be addressed to Per-Olaf Hasselgren, MD, Department of Surgery, University of Cincinnati, 231 Bethesda Avenue, Ciinnati, Ohio 45267-0558. Presented at the 32nd Annual Meeting of the Society for Surgery of the Alimentary Tract, New Orleans, Louisiana, May 20-22,199l.
on
AND METHODS
Studies were performed in male Sprague-Dawley rats (Zivic-Miller, Zelienople, PA) weighing 150 to 200 g. Rats were housed two per cage at 22OC in a room with a 12-hour light/dark cycle. Two series of experiments were performed. In the first series of experiments, the effect of sepsis on plasma levels of gut peptides was studied. Sepsis was induced by cecal ligation and puncture (CLP) as described previously 1121. Control animals underwent sham operations. Rats were resuscitated with normal saline (5 mL/ 100 g body weight) administered subcutaneously on the back at the time of operation. To avoid any influence of different food intake on metabolic changes, both groups of rats were fasted but had free access to water following the operative procedures. Sixteen hours after CLP or sham operation, animals were anesthetized with ether and 3 to 4 mL of blood was drawn from the right atrium (referred to as systemic blood) and the portal vein. The blood was collected into chilled heparinized
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
181
TABLE I Portal and Systemic Plasma Concentrations @g/ml) of Gut Peptides in Control (n = 7) and Septic (n = 9) Rats Systemic
VIP PW Secretin Substance P Gastrin GRP
Portal
Sham
CLP
3.0 -+ 1.9 32.1 2 5.0 23.5 lr 4.4 ND ND ND
40.7 + 13.4* 73.4 + a.o* 95.3 f 14.7* ND ND ND
CLP
Sham 16.4 46.3 30.1 66.6 0.96 16.3
2 + + + 2 2
4.6f 4.6 5.2 13.7 0.44 3.2
213.8 ” 46.2*+ 108.4 k i7.2*+ 127.2 100.1 14.6 14.7
2 13.5* * 13.7 f 2.6* + i .a
CLP = cacal ligation and puncture; VIP = vasoactive intestinal peptfde; PYY = peptide w; GRP = gaatrin-releasingpeptide; ND = not determined. *p < 0.05 varsua sham. tp ~0.05 varaus systemic blood in the corresponding group.
TABLE II Portal and Systemic Plasma Concentrations (pglml) of Gut Peptides In Control (n = 6 to 10) and TNF-Treated (n = 7 to 10) Rats Portal Vein
VIP PW Secretin
Systemic Blood
Control
TNF
Control
TNF
42.3 + 4.5 52.6 2 6.7 67.0 + 9.6
53.7 ? 11.7 156.6 f 26.2* 75.7 f 15.0
24.4 -+ 3.5 66.4~ 10.1 69.0 + 6.5
26.6 -t 5.9 99.1 2 10.1 69.5 f 21.2
TNF = tumor necrosis factor; VIP = vasoactiva intestinal peptide; PW = peptide ‘r-f. lp co.05 versus control.
TABLE III Portal Plasma Concentrations @g/ml) of Gut Peptides in Control (n = 7 or 6) and IL-l-Treated Rats (n = 6 to 9) Control VIP PW Secretin Substance P Gastrin GRP
95.1 103.4 41.4 33.6 12.5 16.7
2 5.6 + 19.5 2 6.0 -+ 6.1 * 3.9 + 2.4
IL-1 76.1 134.1 23.5 31.6 21.1 12.4
+ t f + 2 +
5.7 15.2 3.0 4.6 4.1 1.6
IL-1 = Interleukin-1; VIP = vasoactive intestinal peptide; PW = peptide W; GRP = gastrin-releasingpeptide.
glass tubes containing aprotinin (400 KIU/mL blood, Sigma Inc., St. Louis, MO) and centrifuged at 3,500g for 15 minutes at 4%. Plasma was stored at -7O’C until assayed (see below). Animals were studied 16 hours after induction of sepsis in the current report because, in previous studies, changes in protein metabolism, including increased intestinal protein synthesis, were observed at this time point [I ,I 21. In the second series of experiments, the effect of TNF or IL-1 on plasma levels of gut peptides was examined. Human recombinant TNFa (rTNFar) was kindly provided by Knoll Pharmaceuticals, Whippany, NJ, and had a 182
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
specific activity of more than 5 X lo6 U/mg protein (L929 cell cytotoxicity) and an endotoxin concentration of less than 5 EU/mg protein (lymulus lysate assay). The cytokine was dissolved in 5 mM phosphate-buffered saline (PBS), pH 7.4, to a final concentration of 10 pg/mL. Human recombinant IL- la (rIL- 1a) was kindly provided by Hoffmann-La Roche, Inc., Nutley, NJ, and had a specific activity of 3 X lo* U/mg protein (D-10 assay) and an endotoxin concentration leas than 0.75 EU/mg protein (lymulus lysate assay). The cytokine was dissolved in PBS to a final concentration of 10 pg/mL. Recombinant TNFa or rIL- 1a! was injected intraperitoneally iKith a total dose of 300 pg/kg body weight, divided into three equal doses with 8-hour intervals. Control animals received corresponding volumes of PBS. This protocol of TNF or IL-l treatment resulted in metabolic changes in skeletal muscle in previous experiments in our laboratory [ 13,14], Rats had free access to water but food was withheld after the first cytokine or control injection in order to avoid any influence of reduced food intake on metabolic changes caused by the cytokines. Two hours after the last cvtokine or control injection, animals were anesthetized with ether, and blood samples were drawn and handled as described above for septic rats. Plasma levels of VIP, substance P, gastrin, GRP, and secretin were determined by radioimmunoassay (RIA), using kits from Peninsula Laboratories (Belmont, CA). RIA of PYY was performed as described previously [ 151, using synthetic porcine PYY labeled with Na1251by the chloramine-T method and antiserum raised in rabbits immunized against porcine PYY [16]. The minimal level of PYY detected with this technique is 20 pg/mL. Results are expressed as mean f SEM. Student’s ttest or analysis of variance followed by Duncan’s test was used for the statistical analysis. RESULTS In the first series of experiments, the mortality rate 16 hours after CLP was 25% (6 of 24). No control rats died. All rats surviving CLP showed signs of sepsis including lethargy, piloerection, diarrhea, and discharge around the nostrils. Basal plasma concentrations of VIP were higher in portal than in systemic blood (Table I). Sepsis resulted 163
JANUARY
1992
EFFECT OF SEPSIS ON GUT PEPTIDE
in increased plasma levels of VIP, PYY, secretin, and gastrin, with the highest concentrations found in portal blood (Table I). Sepsis did not significantly affect portal plasma concentrations of substance P or GRP, and these peptides were not measured in systemic blood. In the second series of experiments, there was no mortality following either TNF or IL-1 injections, but the cytokinetreated animals showed moderate lethargy, piloerection, and diarrhea. Administration of TNF resulted in increased plasma levels of PYY in portal blood but no other significant changes were noted in VIP, PYY, or secretin levels (Table II). Substance P, gastrin, and GRP were not measured in this experiment. In the experiment in which IL-l was administered, only portal vein plasma concentrations of the gut peptides were measured. IL-l did not significantly affect plasma levels of any of the peptides measured (Table RI). It should be noted that control plasma levels of the gut peptides were generally lower in the first series of experiments (Table I) than in the second series of experiments (Tables II and III), both in portal and systemic blood. This may reflect the fact that control rats underwent different procedures in the different experimental series and that the experiments were performed at different times. It is important to point out, however, that control and experimental groups were always studied at the same time within the different experiments. COMMENTS
The current study showed that sepsis in rats resulted in increased plasma levels of the gut peptides VIP, PYY, secretin, and gastrin. Thus, the response to sepsis does not seem to be specific for VIP, which was already shown in previous reports to be increased during sepsis [4] or endotoxemia [2,3]. On the other hand, sepsis probably does not stimulate the release of all gastrointestinal peptides because, in the current study, plasma levels of substance P and GRP were not affected by sepsis. Although increased plasma levels of gut peptides during sepsis, in particular in portal blood, may be consistent with increased intestinal release into the circulation, the results may also reflect reduced clearance of the peptides, especially in view of their short half-life times of a few minutes only. It should also be noted that a reduction in portal blood flow could give rise to increased portal concentrations of the peptides even if production rates were unchanged. Therefore, measurements of portal blood flow would have been necessary to calculate actual release rates of the peptides. Previous studies from our laboratory demonstrated that rats are hemodynamically stable 16 hours after CLP [Z7]. In other reports, intestinal blood flow in rats was increased during septic peritonitis [18] or unchanged during endotoxemia [19]. Thus, it is unlikely that the current results of increased portal plasma concentrations of the peptides reflected decreased blood flow. For the same reason, it is unlikely that gut ischemia was responsible for increased release of the pep tides as suggested previously for VIP [2U]. Increased release of gut peptides during sepsis can result from increased synthesis of the peptides and/or
stimulated release of preexisting intracellular pools. In recent experiments in our laboratory, intestinal protein synthesis was increased in septic rats with no concomitant change of protein content in the intestinal wall [I], a result that may be consistent with stimulated synthesis of secreted proteins. Thus, increased release of gut peptides during sepsis may at least partly reflect stimulated synthesis of the substances. The current results suggest that the release of gut peptides during sepsis is not primarily mediated by TNF or IL- 1, although TNF may be partly responsible for the release of PYY. It is important to point out, however, that the results do not rule out the possibility that TNF and IL-1 regulate the release of gastrointestinal hormones during sepsis. Cytokines may have different biologic effects in the septic organism than in healthy rats since various factors present during sepsis may act as important co-factors to the cytokines. It is also possible that the amounts of cytokines administered here were not large enough to induce sepsis-like plasma or tissue concentrations of the cytokines. It should be noted, however, that the same amounts of TNF or IL- 1 as used here resulted in metabolic alterations in skeletal muscle that were almost identical to those observed during sepsis [13,14]. Further experiments, utilizing specific receptor blockers or cytokine antibodies, are necessary to test the hypothesis that cytokines regulate the release of gastrointestinal hormones during sepsis. The biologic significance of elevated plasma levels of gut peptides during sepsis is unknown and can only be speculated on at present. Some of the physiologic alterations induced by sepsis may be mediated or affected by gut peptides. For example, gastric erosions or ulcerations during sepsis may be caused by gastrin-induced acid secretion [21]. Gut peptides have been implicated in delayed gastric emptying [22,23] and impaired intestinal motility [24-24 and may play a role in the pathogenesis of motility abnormalities induced by sepsis. Some metabolic changes characteristic of sepsis, such as glycogenolysis and lipolysis, can be elicited by VIP [27,28]. This peptide may also affect the hemodynamic changes induced by sepsis through its vasodilatory action [29] and positive ionotropic [30] effects. Other peptides may have a beneficial effect during sepsis by enhancing intestinal mucosal growth [31,32]. In summary, the current results suggest that sepsis induces increased plasma levels of most, but not all, gastrointestinal hormones. Except for PYY, a similar response was not found after TNF or IL-l administration. Further studies are necessary to define the mechanisms and biologic implications of gut peptide release in sepsis. REFERENCES 1. Von Allmen D, Hasselgren PO, Higashiguchi T, Zamir 0, Fischer JE. Increased intestinal protein synthesis during sepsis. Surg Forum 1990; 41: 68-70. 2. Freund H, Ebeid AM, Fischer JE. An increase in vasoactive intestinal peptide levels in canine endotoxin shock. Surg Gynecol Obstet 1981; 152: 604-6. 3. Revhaugh A, Lygren I, Lundgren TI, Andersen OK, Burhol PG,
THE AMERICAN JOURNAL OF SURGERY
VOLUME 163 JANUARY 1992
183
23. Valenzuela JE, Defilippi C. Inhibition of gastric emptying in humans by secretin, the cctapeptide of cholecystokinin, and intraduodenal fat. Gastroenterology 1981; 81: 898-902. 24. Mukhopadhyay AK, Johnson LR, Copeland EM, Weisbrodt NW. Effect of secretin on electrical activity of the small intestine. Am J Physiol 1975; 229: 484-8. 25. Kachelhoffer J, Mendel C, Dauchel J, Hohmatter D, Grenier JF. The effect of VIP on intestinal motility. Study on ex uiuo perfused isolated canine jejunal loops. Am J Dig Dis 1976; 21: 957-62. 26. Krantis A, Potwin W, Harding RK. Peptide YY (PYY) stimulates intrinsic enteric motor neurons in the rat small intestine. Arch Pharm (Weinheim) 1988; 338: 287-92. 27. Kevins C, Said SI. Hyperglycemic and glycogenolytic effects of vasoactive intestinal polypeptide. Proc Sot Exp Biol Med 1973; 142: 1044-7. 28. Frandsen EK, Moody AJ. Lipolytic action of newly isolated vasoactive intestinal polypeptide. Horm Metab Res 1973; 5: 196-9. 29. Said SI, Mutt V. Potent peripheral and splanchnic vasodilator peptide from normal gut. Nature 1970; 225: 863-4. 30. Said SI, Mutt V. Polypeptide with broad biological activity: isolation from small intestine. Science 1970; 169: 1217-8. 31. Bloom SR. Gut hormones in adaptation. Gut 1987; 28: 31-5. 32. Johnson LR, Lichtenberger LM, Copeland EM, Dudrick SJ, Castro GA. Action of gastrin on gastrointestinal structure and function. Gastrcenterology 1975; 68: 1184-92.
Giercksky KE. Increased plasma levels of vasoactive intestinal polypeptide in pigs during endotoxemia. Em Surg Res 1985; 17: 75-82. 4. Fuortes M, Blank MA, Scalea TM, Pollock TW, Jaffe BM. Release of vasoactive intestinal peptide during hyperdynamic sepsis in dogs. Surgery 1988; 104: 894-8. 5. Weiner I, Khalil T, Thompson JC, Rayford PL. Gastrin. In: Thompson JC, Greely GH, Rayford PL, Townsend CM, editors. Gastrointestinal endocrinology. New York: McGraw Hill, 1987; 194-212. 6. Besson J, Laburthe M, Bataille D, Dupont C, Rosselin G. Vasoactive intestinal peptide (VIP): tissue distribution in the rat as measured by radioimmunoassay and by radioreceptorassay. Acta Endccrinol 1978; 87: 799-810. 7. Straus E, Yalow RS. Immunoreactive secretin in gastrointestinal muccsa of several mammalian species. Gastroenterology 1978; 75: 401-4. 8. Lundberg SM, Tatemoto K, Terenius L, et al. Localization of peptide YY (PYY) in gastrointestinal endocrine cells and effects on intestinal blood flow and motility. Proc Nat1 Acad Sci U S A 1982; 79: 4471-5. 9. Greeley GH, Spannagel A, Burdet JB. Distribution of gastrinreleasing peptide and bombesin-like peptides in the alimentary canal of rats, rabbits, dogs and humans [abstract]. Gastroenterology 1984; 86: 1097. 10. Holzer P, Bucsics A, Saria A, Lembeck F. A study of the concentration of substance P and neurotensin in the gastrointestinal tract of various mammals. Neuroscience 1982; 7: 2919-24. 11. Flores EA, Bistrian BR, Pomposelli JJ, et al. Infusion of tumor necrosis factor/cache&in promotes muscle catabolism in the rat. A synergistic effect with interleukin-1. J Clin Invest 1989; 83: 1614-22. 12. Hasselgren PO, Talamini MA, James JH, et al. Protein metab olism in different types of skeletal muscle during early and late sepsis in rata. Arch Surg 1986; 121: 918-23. 13. Zamir 0, Hasselgren PO, von Allmen D, Fischer JE. The effect of interleukin-1 and the glucocorticoid receptor blocker RU 38486 on total and myotibrillar protein breakdown in skeletal muscle. J Surg Res 1991; 50: 579-83. 14. Zamir 0, Hasselgren PO, Kunkel SL, Higashiguchi T, Frederick JA, Fischer JE. Evidence that TNF participates in the regulation of muscle proteolysis during sepsis. Arch Surg. In press. 15. Chen MH, Balasubramaniam A, Murphy RF, et al. Sensitive radioimmunoassay for measurement of circulating peptide YY. Gastroenterology 1984; 87: 1332-8. 16. Balasubramaniam A, O’Hare MMT, Murphy RF, Joffe SN, Buchanan KD, Fischer JE. Synthesis of PYY for production of antibodies [abstract]. Regul Pept 1984; 7: 275. 17. Pedersen PV, Warner BW, Bjornson HS, et al. Hemodynamic and metabolic alterations during experimental sepsis in young and adult rats. Surg Gynecol Obstet 1989; 168: 148-56. 18. Lang CH, Bagby GJ, Ferguson JL, Spitzer JJ. Cardiac output and redistribution of organ blood flow in hypermetabolic sepsis. Am J Physiol 1984; 246: R331-7. 19. Jepson MM, Cox M, Bates PC, et al. Regional blood flow and skeletal muscle energy status in endotoxemic rata. Am J Physiol 1987; 252: E581-7. 20. Modlin IM, Bloom SR, Mitchell S. Plasma vasoactive polypeptide (VIP) levels and intestinal ischaemia. Experientia 1978; 34: 535-6. 21. Gdonkor P, Mowat C, Himal HS. Prevention of sepsis induced gastric lesions in dogs by cimetidine via inhibition of gastric secretion and by prostaglandin via cytoprotection. Gastroenterology 1981; 80: 375-9. 22. Debas HT, Farooq D, Grossman MI. Inhibition of gastric emptying is a physiological action of cholecystokinin. Gastroenterology 1975; 68: 1211-7.
When you give these cytokines intraperitoneally, how are they absorbed? Are they absorbed quickly or over a period of several hours? How is intraperitoneal administration compared with intravenous injection? Haile T. Dehas (San Francisco, CA): I noticed that the levels of the two neuropeptides, substance P and gastrin-releasing peptide, were not raised. Is it possible that there was actually no increase in release, but that there was a reduction in mesenteric blood flow? Did you measure portal blood flow? William Silen (Boston, MA): My question has to do with whether you determined if this was simply a release of stored gastrin rather than stimulation of formation. In other words, was there destruction or injury of the intestinal mucosa or other portions of the gastrointestinal tract that might have stored peptides that would be released into the portal circulation during shock? Mark Evers (Galveston, TX): Have you measured actual peptide YY content? Are you increasing new peptide synthesis or simply stimulating release in response to sepsis? Oded Zamir (closing): Selection of the doses of recombinant interleukin- 1 (rIL- 1(Y)and recombinant tumor necrosis factor (rTNF-ol) was based on previous studies in our laboratory, which showed changes in protein metabolism in skeletal muscle and gut after administration of the cytokines in the same dose. It should be noted, however, that it is unknown if the circulating, and more importantly, tissue levels of the cytokines following injection at this dose are similar to those seen in sepsis or endotoxemia. Absorption of the cytokines following intraperitoneal injection, as reflected by the metabolic and endocrine
184
163
THE AMERICAN
JOURNAL
OF SURGERY
VOLUME
DISCUSSION Dr. Cheung:
JANUARY
1992
EFFECT OF SEPSIS ON GUT PEPTIDE
response, is very rapid. For example, we found a striking elevation of plasma corticosterone as early as 30 minutes after intraperitoneal injection of either rIL-la or rTNF-a. Dr. Debas, we have not measured portal or mesenteric blood flow in the current study. However, other investigators found mesenteric blood flow to be unchanged or increased after peritonitis or endotoxemia in rats. Thus, elevated portal levels probably reflect increased release of the peptides rather than higher concentrations due to reduced blood flow after sepsis.
THE AMERICAN
Dr. Silen, this is an important question but, from the current data, we cannot exclude the possibility that elevated plasma levels of the peptides reflect release of stored peptides rather than increased synthesis. Dr. Evers, we did not measure tissue content of gut peptides. However, the fact that plasma levels of these very short-lived peptides were elevated 16 hours after induction of sepsis supports the idea that elevated plasma levels reflect increased synthesis in the gut; further studies are necessary to address this question.
JOURNAL
OF SURGERY
VOLUME
163
JANUARY
1992
185
