GASTROENTEROLOGY

1992;103:1437-1443

Endotoxin-Induced Alterations in Rat Colonic Water and Electrolyte Transport MAE J. CIANCIO, University

LUIGI

of Chicago, Department

VITIRITTI, of Medicine,

AAMI

This study examines the effects of endotoxin on intestinal water and electrolyte transport in adult male rats. Endotoxin (1.55 mg/kg, intravenously) reduced in vivo colonic saline absorption 61% in 1 hour. In vitro unidirectional and net “Na and 3sC1 fluxes showed that endotoxin significantly decreased net colonic 22Na absorption compared with control colons (0.3+ 1.7vs.4.6f 1.1pEq/h X cm2). Although endotoxin had no significant effect on basal short circuit current (IX) and conductance, 3Hinulin flux studies suggested an increase in colonic permeability. Isc responses to the 5’-cyclic adenosine monophosphate (CAMP)-dependent secretagogues prostaglandin E, (1 pmol/L) and vasoactive intestinal peptide (0.1 pmol/L) were diminished by 80% and 50%, respectively. However, cytosolic CAMP-dependent protein kinase activity under basal and stimulated (6pmol/L 8-bromo-CAMP) conditions was not altered by endotoxin treatment. The Isc responses to 10 pmol/L bethanechol, a Ca2+-dependent agonist, were not effected by endotoxin treatment. It was concluded that endotoxin significantly affects colonic transport function and may contribute to the development of diarrhea in inflammatory bowel diseases.

T

DHAR,

and EUGENE

B. CHANG



Chicago, Illinois

he diarrhea associated with inflammatory bowel diseases (IBD) is a frequent and debilitating complication. Inflammatory mediators such as prostaglandins,‘-3 leukotrienes,2*4 oxygen derived free radicals,’ and platelet activating facto? have been implicated as having major roles in stimulating intestinal secretion and compromising mucosal integrity. However, the factors that initiate their production are poorly understood. In this study, we examined the possibility that endotoxin, a lipopolysaccharide component of gram-negative bacterial cell wall, may be a causative agent in stimulating colonic secretion and altering epithelial cell function. Endotoxin has been shown to be a potent activator of the inflammatory process, stimulating the production and release of numerous cytokine$ by activated macrophages and other immune cells. In IBD, where

intestinal permeability may be compromised inherently as a consequence of active disease,7 intraluminal endotoxin may permeate the colonic epithelial barrier, activate mucosal immune cells, and circulate systemically. The latter, for instance, has been documented in patients with active Crohn’s disease’,’ and ulcerative colitis.g In previous studies where endotoxin has been administered intravenously (IV), significant dose-related gastrointestinal changes, ranging from mild diarrhea accompanied by focal areas of hyperemia and inflammation to bloody, watery diarrhea accompanied by extensive necrosis and bleeding, have been observed.“*” However, the pathophysiological basis of these phenomena, particularly in reference to the development of diarrhea, remain poorly defined. Therefore, the purpose of this study is to evaluate the effects of endotoxin on in vivo intestinal saline absorption and to define the basis of any observed effects by measuring in vitro electrolyte transport and mucosal permeability under basal and secretagogue-stimulated conditions. Materials and Methods Animals Adult male Harlan Sprague-Dawley rats (350-40~1 g) were fed standard rat chow and water ad libitum before experimentation. Rats were randomly assigned to one of two treatment groups: (a) control rats given an IV bolus injection of 0.9% NaCl or (b) endotoxin-treated rats given an IV bolus injection of Salmonella enteritidis lipopolysaccharide B lot no. 774541 (1.55 mg/kg; Difco Laboratories, Detroit, MI). All experiments adhered to the National Institutes of Health guidelines for the use of experimental animals. Stool Measurements Each control and endotoxin-treated rat was placed into a separate cage, the bottom of which was covered with filter paper and wire mesh to allow for the collection of all fecal material excreted during the 1 hour after injection. 0 1992 by the American

Gastroenterological 0016-5065/92/$3.00

Association

1438

CIANCIO ET AL.

GASTROENTEROLOGY Vol. 103, No. 5

At the end of the time period, each rat was removed from its cage, and the fecal pellets were counted and weighed. Intestinal

Loops

Under ether anesthesia, intestinal loops were made in the jejunum, ileum, and colon of control and endotoxintreated rats as previously described.” On completion of surgery, while still under ether anesthesia, each rat was injected IV with either saline or 1.55 mg/kg endotoxin and allowed to recover. After 1 hour, the rats were killed with ether and the loops removed. Each loop was weighed, its length determined, emptied, reweighed, and intestinal saline absorption determined as previously described.” Ussing

Chamber

determined in the same tissues using **Na (2 pCi) and 3sCl (10 pCi) in 10 mL of KRB. Fluxes were measured as previously described.13 For the inulin flux experiments, 4 sequential 15-minute flux periods were determined. For the “Na and 36Cl flux experiments, a 30-minute control flux period was followed by a 20-minute flux period with 0.1 pmol/L VIP. 22Na and 36Cl fluxes were quantitated as previously described.13 Briefly, samples were initially counted to determine the gamma counts attributed to “Na and then recounted to determine the beta counts attributed to 36Cl. Because of the crossover of “Na counts in the 36Cl counts, all 36Cl counts were corrected for the interfering **Na counts by counting known amounts of *‘Na and determining the corresponding counts that occur with the beta counter.

Experiments

One hour after the IV injection of saline or 1.55 mg/kg endotoxin to control and endotoxin-treated rats, respectively, the rats were euthanized with ether and their distal colons removed. As previously described,13 the distal colon was removed, was placed immediately in ice cold Krebs bicarbonate buffer (KRB) containing (in mmol/ L) 114 NaCl, 5 KCl, 1.65 Na,HP04, 0.3 NaH,PO,, 1.25 CaCl,, 1.1 MgCl,, and 25 NaHCO,, and was gently rinsed free of its fecal contents. The muscle layer was removed by blunt dissection and approximately l-cm segments of distal colon were mounted in Ussing chambers as previously described.13 The tissues were bathed with KRB (37’C, pH = 7.4) containing 10 mmol/L glucose. The bathing solution was continuously gassed with water saturated 95% 0,/5% co,. Transmural potential difference, short-circuit current (Isc), and conductance (G) were measured as previously described.13 All tissues were maintained under short circuited conditions except for the brief period when PDs were recorded. Thirty minutes after the tissue stabilized or at the completion of a flux experiment, changes in Isc were determined in response to stimulation with prostaglandin E, (PGE,; 1 pmol/L), vasoactive intestinal peptide (VIP; 0.1 pmol/L), forskolin (100 pmol/L), 8-bromo-3’,5’cyclic adenosine monophosphate (1 mmol/L; 8-bromoCAMP), theophylline plus 8-bromo-CAMP (10 mmol/L and 1 mmol/L, respectively), and carbamyl B-methylcholine chloride (10 pmol/L; bethanechol). All stimulating agents were added to the serosal side of the tissue except for theophylline, which was added to both the mucosal and serosal sides.

CAMP-Dependent Determination

Protein

Kinase

CAMP-dependent protein kinase (A kinase) activity was determined using a modified histone phosphorylation assay.” Briefly, distal colons were removed from control and endotoxin-treated rats, were stripped of the muscle layer, and were homogenized in ice cold 10 mmol/L Tris containing 1 mmol/L ethylenediamine tetraacetic acid, 1 mmol/L ethyleneglycol-bis(Paminoethy1 ether)-N,N’-tetraacetic acid, 0.05% P-mercaptoethanol, 1 mmol/L phenylmethyl sulfonyl fluoride and 10 pg/mL leupeptin. The homogenates were spun at 2000 g X 10 minutes followed by a second spin at 100,000 g X 1 hour. Cytosolic protein (20-30 pg) was added to duplicate tubes to determine the phosphorylation of histone II-AS with 2 PCi of 32P-adenosine triphosphate (ATP) under basal and stimulated (6 pmol/L 8-bromo-CAMP) conditions in the absence and presence of the A kinase inhibitor PKI (Gibco/BRL, Grand Island, NY; 1 pmol/L). After 10 minutes at 3O”C, 50 pL of sample was immediately blotted onto Whatman 3-mm filter paper squares (Whatman, Hillsboro, OR) to stop the reaction, air dried, repeatedly washed in ice cold 10% trichloroacetic acid (TCA) to remove unincorporated 32P, and washed in 95% ethanol followed by a final wash in ethyl ether. The filter papers were air dried, placed in liquid scintillation vials, and counted for 10 min/sample. Statistics Comparisons were made by Student’s t test for paired and unpaired data. Values are expressed as means t SEM. Significance is reported for P I 0.05.

Flux Measurements Unidirectional mucosal to serosal (m-s) and serosal to mucosal (s-m) fluxes (J) were determined for tissue segments paired by resistances that differed by less than 25%. Unidirectional fluxes were measured beginning 20-30 minutes after mounting the tissue and 15-25 minutes after adding the radioactive agent. For the inulin flux experiments, 3H-inulin (10 pCi; Amersham, Arlington Heights, IL) was added to KRB containing 1 mmol/L inulin. To assess mucosal permeability, only s-m 3H-inulin fluxes were determined as described by Madara and Dharmsathaporn.14 Sodium and chloride fluxes were simultaneously

Results Rats were injected IV with 1.55 mg/kg of endotoxin, a dose specifically selected because it reproducibly induces diarrhea within 1 hour in fed rats and yet is nonlethal during this time course.‘” Diarrhea was assessed by weighing the fecal contents expressed from each control and endotoxin-treated rat for l-hour postinjection. Endotoxin treatment significantly increased (P I 0.05) the weight and frequency of fecal material released; the weight of fecal mate-

November

1992

ENDOTOXIN

0

120 1 I

Control Endotoxin

-

I

100 80 .60 40 20 0 JEJUNUM

ILEUM

COLON

Figure 1. Effects of endotoxin on intestinal saline absorption. Jejunal, ileal, and colonic loops were made in control (n = 11) and endotoxin-treated (n = 11) rats as described in Materials and Methods. The results are expressed as the microliters of saline absorbed per centimeter of intestinal loop (mean f SEM). *P I 0.05control vs. endotoxin treatment.

rial released from endotoxin-treated and control rats was 1.85 + 0.65 g (n = 6)and 0 f 0 g (n = 7),respectively, and the frequency of stool released was 3.8+ 1.4and 0 f 0 stools/h for each respective group. The fecal contents expressed from the endotoxin-treated rats were initially well-formed pellets that gradually became more watery and less formed during the lhour time course. To localize the site of endotoxin’s action, saline absorption was determined in jejunal, ileal, and coionic loops, As shown in Figure 1, endotoxin significantly reduced saline absorption in the distal colon loops within 1 hour but had no effect on jejunal or ileal loop absorption. Colonic saline absorption was reduced by 61% in endotoxin-treated rats compared with saline-injected controls (P I 0.05). For the remainder of this study, mucosal strips of colonic tissues isolated from control and endotoxin-treated rats were mounted in Ussing chambers to examine the potential mechanisms of action of endotoxin on fluid and electrolyte transport. As shown in Table 1, endotoxin did not significantly affect basal short circuit current (1s~). The average basal Isc’s were 2.24 ?Z 0.3 and 2.06 f 0.1 nEq/h X cm2 for the control (n = 17) and endotoxintreated (n = 21) colons, respectively. However, the in vivo administration of 1.55 mg/kg of endotoxin appeared to significantly reduce basal net “Na absorption compared with controls; net “Na absorption was 4.8+ 1.1yEq/h X cm2 in control colons (n = 8) and 0.3k 1.7nEq/h X cm’ in endotoxin-treated colons (n = 10). This change in net “Na absorption ap-

AND INTESTINAL

TRANSPORT

1439

peared to result largely from an elevation in the unidirectional J,_, for “Na, approaching values equivalent to J,_,. However, endotoxin treatment did not significantly alter net 36C1absorption. A previous study by Edmonds and Pilcher17 showed an increase in plasma to lumen flux rates of sodium in rectal mucosa of patients with ulcerative colitis, suggesting an increase in colonic permeability. Therefore, we examined the possibility that endotoxin significantly reduced net 22Na absorption by increasing colonic permeability. To assess this possibility, 3H-inulin JS_m,a commonly used indicator of tissue permeability,14 was measured in control and endotoxin-treated colonic tissues. As shown in Figure 2, fluxes during all four time periods were elevated in the endotoxin-treated colons, reaching statistical significance by 60-80 minutes. However, tissue conductances were not statistically different in endotoxin-treated tissues: 16.8 + 1.4 millimhos/ cm2 (n = 21) in the endotoxin-treated colons and 14.4 + 1.0 millimhos/cm2 (n = 17) for the control colons (Table 1). The lack of a significant change in tissue conductance in the endotoxin-treated tissues suggested that endotoxin induced subtle, focal changes in tissue permeability as reflected in the elevated inulin fluxes. This is supported by our histological data (not shown) that showed no major structural alteration in the endotoxin-treated colons compared with controls. We next examined the functional response of coionic tissues from control and endotoxin-treated rats

20-35

35-50

SO-65

65-80

TIME (minutes) Figure 2. Effects of endotoxin on 3H-inulin flux. Unidirectional Js-m of ‘H-inulin are reported for colonic tissues from control (n = 10) and endotoxin-treated rats (n = 8) as described in Materials and Methods. The values reported are pmols of 3H-inulin crossing the mucosal tissue per hour x cm* (mean f SEM). *P I 0.05control vs. endotoxin treatment.

GASTROENTEROLOGY Vol. 103,No. 5

1440 CIANCIO ET AL.

Table 1. Effects of Endotoxin N+,-,

on Sodium and Chloride Fluxes Under Basal and VIP-Stimulated Na,,.,

Wnet

q,.,

Chll-,

qnet

Conditions Isc

G

Control A. Basal B. VIP (B-A) Endotoxin A. Basal B. VIP (B-A)

13.8f 1.5 12.1f 1.3 -1.6f 0.7

8.9+ 0.7 9.3+ 1.0 0.4rfr 0.4

4.8+ 1.1 2.8f 0.6' -2.0* 0.7

21.0f 2.6 13.2f 1.4' -7.8+ 2.8

15.9f 1.6 18.8+ 4.5 2.9i 3.5

5.1f 1.6 -5.5+ 3.8 -10.6i 4.8

12.6f 0.8 12.6+ 1.1 0 + 0.8

12.3f 1.8 12.7f 1.8 0.4+ 0.5

0.3f 1.7b -0.1f 2.1 -0.4f 1.0

19.9f 1.6 19.0f 1.8b -0.9+ 2.5

18.3+ 2.4 19.1+ 3.9 0.8k 2.6

1.6k 2.3 -0.1+ 3.4 -1.7+ 2.3

2.24rt0.3 5.72+ 0.4a 3.48f 0.3

14.4k 1.0 17.3t 1.3O 2.9* 0.3

2.06f 0.1 16.8k 1.4 3.23f 0.2"lb 18.4+ 1.4' 1.17+ 0.2 1.6f 0.2

NOTE. The flux values for “Na and 36C1,as well as the Isc are expressed as the yEq/h X cm’ (mean + SEM) for colonic mucosa isolated rats (1.55 mg/kg IV; n = 10). The basal flux period was 30 minutes followed by a 20-minute from control (n = 7-8)and endotoxin-treated stimulated flux period with 0.1 pmol/L VIP. Isc values and conductances (G; mmhos/cm’) are the means of the average values determined during each flux period for control (n = 17) and endotoxin-treated (n = 21)tissues. The (B-A)values represent the difference between the basal and VIP-stimulated values (mean + SEMI. Jm-s, mucosal to serosal flux: Js-m, serosal to mucosal flux; Jnet = J,.,-J,.,. “P I 0.05 basal vs. VIP. bP I 0.05 control vs. endotoxin.

to stimulation with the CAMP-mediated secretagogues PGE, and VIP and the Ca’+-mediated secretagogue bethanechol. As shown in Figure 3, the increases in Isc induced 10 minutes after the addition of 1 pmol/L PGE, and 0.1 pmol/L VIP were reduced 80% and 54%, respectively, in the endotoxin-treated colons compared with controls (Figure 3). In contrast, the Isc response 5 minutes after addition of 10 pmol/L bethanechol was unaltered in the endotoxin-treated colons (Figure 3). These findings sug-

200 ,

1

0

Control Endotoxin

PROSTAGIANDIN E2

VIP

Bl?WANECHOL

Figure 3. Effects of endotoxin on the short circuit current (1s~) response to prostaglandin E,, vasoactive intestinal peptide, and bethanechol. Distal colons from control (n = 20)and endotoxintreated (n = 23)rats were stimulated with 1 pmol/L PGE, (n = 13-15), 0.1pmol/L VIP (n = g-11), or 10 pmol/L bethanechol (n = 5) as described in Materials and Methods. Mean basal Isc’s were 66.4 k 9.2 and 55.0 -C 4.5 pAmps/cm’ for the control and endotoxin-treated tissues, respectively. Values reported are the change in Isc (pAmps/cm’; mean f SEM) 10 minutes after the addition of PGE, or VIP and 5 minutes after the addition of bethanechol. *P 5 0.05control vs. endotoxin-treated tissues.

gested that endotoxin exerts a functional change in stimulated colonic transport by selectively influencing the CAMP-mediated effects and leaving the Ca’+dependent response intact. To establish the ionic basis for the results reported in Figure 3, VIP-stimulated changes in unidirectional 22Na and 3”C1fluxes were measured (Table 1). In normal rat colon, VIP inhibits net sodium absorption and stimulates active chloride secretion.” As shown in Table 1, net 22Na absorption was significantly reduced from 4.8 + 1.1 to 2.8 f 0.6 pEq/hr X cm2 (P 50.05) in control tissues after VIP stimulation. The change appeared to be due to a decrease in unidirectional J,., of sodium (P = 0.06) as well as a modest increase in J,_,. More dramatic, however, were the changes in 36C1transport induced by 0.1 pmol/L VIP. Unidirectional J,_, of 3”C1was significantly reduced (PI 0.05) from 21.0 + 2.6 to 13.2 + 1.4 pEq/hr X cm2 in control tissues. The resultant effect was a reversal from net 36C1absorption (5.1 + 1.6 pEq/hr X cm2) to net 36C1secretion (-5.5 f 3.8 pEq/hr X cm2). In contrast, there was no significant change in 22Na and 36C1 transport in endotoxin-treated tissues stimulated with VIP (Table 1). Unidirectional as well as net 22Na and 3”C1 fluxes remained unchanged from their respective basal values after VIP stimulation. To establish how endotoxin may be acting to limit the colonic response to CAMP-dependent secretagogues, control and endotoxin-treated colonic mucosal strips were stimulated with the following: lOOpmol/L forskolin, an adenylate cyclase stimulator; 1 mmol/L 8-bromo-CAMP, a CAMP analogue; or costimulation with 10 mmol/L theophylline (a phosphodiesterase inhibitor) plus 1 mmol/L 8-bromoCAMP. As shown in Figure 4, the response 10 minutes after stimulation with each of these agents was reduced in the endotoxin-treated colons. Compared with control tissues, endotoxin significantly

November

ENDOTOXIN AND INTESTINAL TRANSPORT

1992

300 1

67 g -2

I

0

Control Endotoxin

FORSKOLIN

1

8-BROMOcAMP TllEOPHYLLlNE + 8-BROMOcAMP

Figure 4. Effects of endotoxin on the Isc response to forskolin, 8-bromo-CAMP, and costimulation with theophylline plus 8bromo-CAMP. Colonic tissue from control (n = 9) and endotoxintreated rats (n = 8) were stimulated with 100 pmol/L forskolin (n = 5) 1 mmol/L 8-bromo-CAMP (n = 5) or costimulated with 1 mmol/L Et-bromo-CAMP plus 10 mmol/L theophylline (n = 6-7) as described in Materials and Methods. Mean basal Isc’s were 118 5 13 pAmps/cm’ and 93 -t 18 pAmps/cm’ for control and endotoxin-treated tissues, respectively. Values reported are the change in Isc (pAmps/cm’; mean + SEM) 10 minutes after the addition of these agents. *P 5 0.05 control vs. endotoxin-treated tissues.

(P I 0.05) reduced the change in Isc response to forskolin from 160 + 31 (n = 5) to 64 -t 21 (n = 5) respectively. Similar responses were uAmps/cm’, observed after stimulation with 8-bromo-CAMP and costimulation with theophylline plus 8-bromoCAMP: the changes in Isc for the endotoxin-treated colons were 32% and 46%, respectively, of the increase elicited in control tissues. These results suggest that the blunted secretory responses to the CAMP-dependent agents, PGE, and VIP, were a result of endotoxin-induced alterations in a step distal to the stimulated increase in CAMP. This hypothesis is supported by our finding that basal CAMP levels were unchanged in endotoxin-treated colons compared with controls: 27.2 k 4.3 pmols/mg protein (n = 6)and 25.7-+1.5pmols/mg protein (n = 5)for endotoxin-treated and control tissues, respectively. Therefore, cytosolic CAMP-dependent protein kinase (A kinase) activities were measured in control and endotoxin-treated rat colons as a possible mechanism for the depressed secretory response to PGE, and VIP. As shown in Figure 5,A kinase activity was determined under basal and stimulated conditions in the absence and presence of the specific A kinase inhibitor PKI (1umol/L). Under all three conditions, the endotoxin-treated colons showed comparable levels of A kinase activity. Moreover, endotoxin did

1441

not significantly change cytosolic A kinase activity determined using control cytosolic A kinase preparations treated with endotoxin in vitro (data not shown). Discussion

This study examined the intestinal electrolyte transport and permeability changes induced after the IV administration of endotoxin, a potent diarrhea-inducing agent.’ This is of particular importance because the colon is normally colonized with bacterial microflora and is a major site of pathology in IBD such as ulcerative colitis. The results from this study suggest that the diarrhea observed 1 hour after the IV administration of 1.55mg/kg of endotoxin resulted from significantly decreased colonic saline absorption and electrolyte transport with no functional changes in small intestinal saline absorption. When closely examined under in vitro conditions, colonic mucosa from endotoxin-treated rats showed a significant decrease in basal net “Na absorption. This change appeared to result from an increase in unidirectional J,_, of “Na, reaching levels comparable to J,.,. A possible explanation for this finding is that endotoxin induced an increase in mucosal permeability, as suggested by the increase in J,., of 3H-

1

BASAL

STIMULATED STIMULA'IEDPLUSPKI

Figure 5. Effects of endotoxin on CAMP-dependent protein kinase activity. Colonic tissues from control (n = 8) and endotoxintreated (n = 8) rats were assayed for CAMP-dependent protein kinase (A kinase) activity as described in Materials and Methods. A kinase activity was assessed under basal and stimulated conditions (6 pmol/L 8-bromo-CAMP) in the absence and presence of the specific A kinase inhibitor, PKI (1pmol/L). Values reported are picomoles of “P transfered/pg of protein X 10 minutes (mean f SEM).

1442

CIANCIO

GASTROENTEROLOGY

ET AL.

inulin in these tissues. Consistent with this notion are previous reports of endotoxin-induced increases in intestinal permeability to intraluminal bacteria and bacterial components.‘Q~20 However, this change in colonic 3H-inulin permeability was not accompanied by a comparable change in tissue conductance (Table 1)because conductivity is a relatively insensitive indicator of mucosal permeability.2* These alterations in basal colonic electrolyte transport and permeability could account for the diarrhea observed in endotoxin-treated animals. In contrast to endotoxin’s effects on basal transport, colons from endotoxin-treated rats showed a decreased secretory response to the secretagogues PGE, and VIP; an effect not entirely consistent with the diarrheagenic actions of endotoxin. Unidirectional J,_, of 22Na and 36C1 remained elevated and there was no increase in J,., of 36C1after stimulation with VIP. The failure of the endotoxin-treated colons to respond to both VIP and PGE, suggested an alteration in the CAMP-dependent second messenger pathway. Contrary to previous reports in cardiac and hepatic tissues,22*23the blunted secretory response to PGE, and VIP did not result from altered colonic CAMP levels. When intracellular CAMP levels were elevated with agents such as forskolin, 8-bromoCAMP and theophylline plus 8-bromo-CAMP, the endotoxin-treated colons still showed a significant reduction in the secretory response compared with controls. This suggested that either specific transport processes were directly affected or that steps distal to CAMP generation were responsible for the decreased secretory response observed in the endotoxintreated colons. In contrast, endotoxin did not alter the secretory capability of colonic mucosa to stimulation with the Ca2+-dependent secretagogue bethanechol. Thus, it appeared that the dysfunctional effects induced by endotoxin were specific for the CAMPdependent pathway and not the result of a global down regulation or alteration in the secretory capacity of the tissue. Our studies suggest that the decreased response of the endotoxin-treated colons results from an alteration in intracellular signaling distal to the generation of A kinase activity because our studies with the specific A kinase inhibitor PKI showed no difference in cytosolic A kinase activity between control and endotoxin-treated colons. The ability of endotoxin to compromise basal coionic water and electrolyte transport as well as increase intestinal permeability’Q~20 suggests a role for luminal endotoxin in the enhanced mucosal “leakiness” associated with inflamed tissues isolated from patients with ulcerative colitis and Crohn’s disease.‘7’24 Endotoxin may be acting directly to influence colonic epithelial cell function or may serve as a signaling agent for the inflammatory response

Vol. 103, No. 5

observed in IBD tissues. In preliminary studies (performed in collaboration with Dr. Gail Hecht, University of Illinois, Department of Medicine/GI), endotoxin treatment for 1 hour had no direct effect on T84 colonic epithelial cell function or permeability. These results suggest that endotoxin may be acting as a triggering agent of other factors in the inflammatory response that may alter mucosal transport function or permeability, either independently or in concert with endotoxin, rather than acting directly on the epithelial cells. Increased production of the cytokines, tumor necrosis factor, and interleukin 1, have been shown in vitro using mucosal biopsies25,2” and mononuclear ce11s27,28from patients with IBD. Other inflammatory agents reported to have a pivotal role in the pathogenesis of IBD, such as prostaglandins,‘-3 leukotrienes,2*4 oxygen derived free radicals,5 and platelet activating factor5 have also been shown to mediate many of the gastrointestinal changes induced by endotoxin.‘0*2Q-31 The correlation between the inflammatory agents produced in IBD and those produced after in vivo endotoxin administration support the concept that luminal endotoxin may be a contributing factor in the pathogenesis of IBD. Unresolved from this study is how or why the in vivo effects of endotoxin were localized to the colon. We speculate that inherent differences may exist in the lamina propria cells of the colon that alter, or their sensitivity to potentially perhaps increase, pathologic agents such as endotoxin. Conceivably, these cells are primed to produce cytokines and other inflammatory mediators by low-grade endotoxin exposure from the colonic microflora, thus making them more responsive to the systematically administered endotoxin in this study. In summary, our data showed several effects of endotoxin on normal colonic mucosa that produce dysfunctional and ion transport changes that may elicit the diarrhea induced by this agent. We speculate that endotoxin may be an important factor in triggering the inflammatory response and the subsequent release of inflammatory mediators in patients with IBD. The end result, if allowed to progress unchecked, is a self-perpetuating process whereby further luminal endotoxin is released, basal water and electrolyte transport is altered, further inflammatory mediator production and release is triggered, and the inflammatory changes in the diseased colon are exacerbated. References 1. Harris DW, Smith PR, Swan CHJ. Determinationofprostaglandin synthetase activity in rectal biopsy material and its significance in colonic disease. Gut 1978;19:875-877. 2. Schumert R, Towner J, Zipser RD. Role of eicosanoids in human and experimental colitis. Dig Dis Sci 1988;33:588-643.

November

1992

3. Sharon P, Ligumsky M, Rachmilewitz D, Zor U. Role of prostaglandins in ulcerative colitis. Gastroenterology 1978;75:638640. 4. Sharon P, Stenson WF. Enhanced synthesis of leukotriene B4 by colonic mucosa in inflammatory bowel disease. Gastroenterology 1984;86:453-460. HM, 5. Bern MJ, Sturbaum CW, Karayalcin SS, Berschneider Wachsman JT, Powell DW. Immune system control of rat and electrolyte transport. J Clin Invest rabbit colonic 1989;83:1810-1820. 6. Morrison DC, Ryan JL. Endotoxins and disease mechanisms. Ann Rev Med 1987;38:417-432. 7. Hollander D, Vadheim CM, Brettholz E, Petersen GM, Delahunty T, Rotter JI. Increased intestinal permeability in patients with Crohn’s disease and their relatives. Annals Int Med 1986;105:883-885. 8. Wellmann W, Fink PC, Benner F, Schmidt FW. Endotoxaemia in active Crohn’s disease. Treatment with whole gut irrigation and 5-aminosalicylic acid. Gut 1986;27:814-820. in ulcerative colitis and 9. Aoki K. A study of endotoxemia Crohn’s disease. I. Clinical study. Acta Med Okayama 1978;32:147-158. 10. Hinshaw LB. Overview of endotoxin shock. In: Cowley RA, Trump BF, ed. Pathophysiology of shock, anoxia, and ischemia. Baltimore: Williams and Wilkins, 1982:219-234. 11. Wallace JL, Steel G, Whittle BJR, Lagente V, Vargaftig B. Evidence for platelet-activating factor as a mediator of endotoxin-induced gastrointestinal damage in the rat. Gastroenterology 1987;93:765-773. 12. Chang EB, Bergenstal RM, Field M. Diarrhea in streptozocintreated rats. J Clin Invest 1985;75:1666-1670. 13. Field M, Fromm D, McCall I. Ion transport in rabbit ileal mucosa. I. Na and Cl fluxes and short-circuit current. Am J Physiol 1971;220:1388-1396. 14. Madara JL, Dharmsathaporn K. Occluding junction structurefunction relationships in a cultured epithelial monolayer. J Cell Biol 1985;101:2124-2133. 15. Roskoski R. Assays of protein kinase. Methods in Enzymology 1983;99:3-6. 16. Ciancio MJ, Hunt J, Jones SB, Filkins JP. Comparative and interactive in vivo effects of tumor necrosis factor alpha and endotoxin. Circ Shock 1991;33:108-120. 17. Edmonds CJ, Pilcher D. Electrical potential difference and sodium and potassium fluxes across rectal mucosa in ulcerative colitis. Gut 1973;14:784-789. 18. Racusen LC, Binder HJ. Alteration of large intestinal electrolyte transport by vasoactive intestinal polypeptide in the rat. Gastroenterology 1977;73:790-796. 19. Deitch EA, Berg R, Specian R. Endotoxin promotes the trans-

ENDOTOXIN AND INTESTINAL TRANSPORT

1443

location of bacteria from the gut. Arch Surg 1987;122:185190. 20. O’Dwyer ST, Michie HR, Ziegler TR, Revhaug A, Smith RJ, Wilmore DW. A single dose of endotoxin increases intestinal permeability in healthy humans. Arch Surg 1988;123:14591464. 21. Marks GJ, Ryan FM, Hidalgo IJ, Smith PL. Mannitol as a marker for intestinal integrity in in vitro absorption studies (abstr). Gastroenterology 1991;100:A697. 22. Tomera JF, Martyn J. Effects of endotoxin infection on cyclic nucleotides in ventricular and gastrocnemius muscle. Circ Shock 1990;32:281-292, 23. Roman0 FD, Jones SB. Beta-adrenergic stimulation of myocardial cyclic AMP in endotoxic rats. Circ Shock 1985;17:243252. 24. Sandle GI, Higgs N, Crowe P, Marsh MN, Venkatesan S, Peters TJ. Cellular basis for defective electrolyte transport in inflamed human colon. Gastroenterology 1990;99:97-105. 25. MacDonald TT, Hutchings P, Choy M-Y, Murch S, Cooke A. Tumour necrosis factor-alpha and interferon-gamma production measured at the single cell level in normal and inflamed human intestine. Clin Exp Immunol 1990;81:301-305. 26. Ligumsky M, Simon PL, Karmelin F, Rachmilewitz D. Role of interleukin 1 in inflammatory bowel disease-enhanced production during active disease. Gut 1990;31:686-689. 27. Satsangi J, Wolstencroft RA, Cason J, Ainley CC, Dumonde DC, Thompson RPH. Interleukin 1 in Crohn’s disease. Clin Exp Immunol1987;67:594-605. 28. Mahida YR, Wu K, Jewel1 DP. Enhanced production of interleukin l-beta by mononuclear cells isolated from mucosa with active ulcerative colitis of Crohn’s disease. Gut 1989;30:835-838. 29. Doherty NS. Inhibition of arachidonic acid release as the mechanism by which glucocorticoids inhibit endotoxin-induced diarrhoea. Br J Pharmac 1981;73:549-554. 30. Flohe L, Giertz H. Endotoxins, arachidonic acid, and superoxide formation. Rev Infect Dis 1987;9:S553-S561. 31. Patton JS, Peters PM, McCabe J, Crase D, Hansen S, Chen AB, Liggitt D. Development of partial tolerance to the gastrointestinal effects of high doses of recombinant tumor necrosis factoralpha in rodents. J Clin Invest 1987;80:1587-1596. Received June 24,1991.Accepted May 20,1992. Address requests for reprints to: Mae J. Ciancio, Ph.D., University of Chicago, Department of Medicine/GI, MC4076,5841 South Maryland, Chicago, Illinois 60637. Supported by National Institutes of Health grant DK38510 and the Digestive Disease Center (DK42086) at the University of Chicago (E.B.C.), a Crohn’s and Colitis Foundation of America grant (E.B.C.) and postdoctoral fellowship (M.J.C.).

Endotoxin-induced alterations in rat colonic water and electrolyte transport.

This study examines the effects of endotoxin on intestinal water and electrolyte transport in adult male rats. Endotoxin (1.55 mg/kg, intravenously) r...
980KB Sizes 0 Downloads 0 Views