THE JOURNAL OF INFECTIOUS DISEASES • VOL. 134, NO.2· © 1976 by the University of Chicago. All rights reserved.
AUGUST 1976
Effect of Enterotoxins of Vibrio cholerae, Escherichia coli, and Shigella dysenteriae Type 1 on Fluid and Electrolyte Transport in the Colon From the Department of Gastroenterology, Walter Reed Army In;titute of Research, Washington, D.C.; and the Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut
Mark Donowltz and Henry J. Binder
Recent studies of the pathogenesis of bacterial diarrhea have identified intestinal secretion of electrolytes in diseases associated with entero-
toxins as well as in diseases for. which no such association has been found [1, 2]. At least seven bacterial enterotoxins have been identified, and, by definition, all induce secretion of fluid by the small intestine [1]. In both cholera and diarrhea associated with noninvasive strains of Escherichia coli, the respective enterotoxins cause secretion of fluid by the small intestine and stimulate adenylate cyclase both in the small intestine and in nonintestinal tissue [1, 2]. Fluid and electrolytes are also secreted during non-enterotoxinassociated bacterial diarrheas such as salmonellosis [3]. In shigellosis, although an enterotoxin has been identified, it is uncertain whether this enterotoxin is involved in the production of diarrhea. In contrast to the extensive evaluation of the effects of the bacterial enterotoxins on the small intestine, it has not been determined whether the colon is involved in the enterotoxin-mediated bacterial diarrheas. Several communications have suggested that Vibrio cholerae enterotoxin does not affect movement of fluid and electrolytes in the colon [4-6]. These observations are surprising in light of the known ability of cholera toxin to increase levels of adenyl ate cyclase in many tissues other than the small intestine (i.e., thyroid, leukocytes, and erythrocytes) [7-10]. In addition, recent observations have shown that adenosine
Received for publication October 16, 1975, and in revised form February 9, 1976. This work was supported by grant no. AM 14669 from the National Institute of Arthritis, Metabolism, and Digestive Diseases, and by grants from the John A. Hartford Foundation, Inc., and the Connecticut Digestive Disease Society. Dr. Donowitz was supported by training grant no. AM 05703 from the National Institute of Arthritis, Metabolism, and Digestive Diseases. We acknowledge the expert technical assistance of J. M. Heffernan, P. Fisher, D. Whiting, and G.Chapo, and the secretarial assistance of Carol Bryan and Elizabeth Penn. We thank Drs. Carl E. Miller (National Institute of Allergy and Infectious Diseases), Thomas M. Jacks (Norwich Pharmacal Company, Norwich, N. Y.), Samuel Formal (Walter Reed Army Institute of Research), and Gerald T. Keusch (Mt. Sinai School of Medicine, New York, N.Y.) for their kind gifts of enterotoxins. In conducting the research described in this paper, the investigators adhered to the "Guide for Laboratory Animal Facilities and Care," as promulgated by the Committee on the Guide for Laboratory Animal Facilities and Care of the Institute of Laboratory Animal Resources, National Academy of Sciences-National Research Council. Please address request for reprints to Dr. Mark Donowitz, Department of Gastroenterology, Walter Reed Army Institute of Research, Washington, D.C. 20012.
135
Downloaded from http://jid.oxfordjournals.org/ at University of Lethbridge on September 6, 2015
Experiments were designed to determine whether the enterotoxins of Vibrio cholerae, Escherichia coli, and Shigella dysenteriae type 1. alter the movement of fluid and electrolytes in the rat cecum. Net secretion of water and sodium were observed after incubation of 1.67 X 10- 5 ltg of purified cholera toxin (choleragen)/ml for 18 hr or of 50 ltg/ ml for 3 hr. The effect of choleragen on cecal transport of water and electrolytes was related to the dose. In addition, choleragen increased cecal mucosal content of adenosine 3': 5'-cyclic phosphate but did not alter the histology of the cecum. The results demonstrate that the colon responds to choleragen in a manner similar to that. of. other tissues. In contrast, the enterotoxins of both E. coli and S. dysenteriae type 1 failed to affect cecal transport of water and electrolytes. These observations may explain several phenomena associated with the diarrhea produced by bacterial enterotoxins.
136
3' :5'-cyclic phosphate (cyclic AMP) induces the secretion of electrolytes in the colon [11, 12], and that guinea pig and rabbit colons bind large amounts of choleragen [13, 14]. The present studies were designed to evaluate whether the enterotoxins of V. cholerae as well as those of E. coli and Shigella dysenteriae type 1 affect transport of fluid and electrolytes by the colon. Materials and Methods
photometry (model no, 143, Instrumentation Laboratories, Woburn, Mass.), and chloride by coulometric titration (Buchler Instrument Division, Fort Lee, N.J.). The concentration of bicarbonate was calculated by measurement of the difference between the concentration of sodium plus potassium and that of chloride. Osmolality was determined by freezing point depression with an osmometer (Advanced Instruments, Needham Heights, Mass.). Enterotoxin preparations. The effect of cholera enterotoxin on the colon was evaluated by inoculation into the cecum of purified cholera enterotoxin (choleragen; lot no. 0972) prepared by Dr. Richard A. Finkelstein (University of Texas Southwestern Medical School, Dallas, Tx.) [16] and generously provided by Dr. Carl E. Miller (National Institutes of Health). The concentration of choleragen, made up in 0.154 M NaCl, ranged from 1.67 X 10- 5 ug/rnl to 166.7 ug/rnl, Heatinactivated (l00 C, 30 min) choleragen (50 ug/rnl) was also studied. The control solution was 0.154 M NaCl. Two different E. coli exterotoxins were studied. (1) A lyophilized, crude, high-molecular-weight enterotoxin prepared from a porcine enteropathogenic strain of E. coli (08 :K87, 88a,b:HI9) [17] was kindly supplied by Dr. T. M. Jacks (Norwich Pharmacal Co., Norwich, N.Y.). This enterotoxin had been shown to contain only the high-molecular-weight, heat-labile form of E. coli enterotoxin; 7.3 mg was the minimal amount necessary to cause a positive loop reaction when incubated for 18 hr in the rabbit ileum. Twice this amount of toxin (5 mg/rnl) , made up in Ringer's bicarbonate solution, was inoculated into rat cecum. (2) Enterotoxin prepared from a porcine enteropathogenic strain of E. coli (08:K87, K88:HI9) was kindly supplied as a whole-cell lysate by Dr. Samuel Formal (Walter Reed Army Institute of Research). This enterotoxin had been shown previously to contain both the heat-labile and heatstable forms of E. coli enterotoxin [18]. The electrolyte concentration of this material, in meq/liter, was: Na, 156; K, 2.8; CI, 146.8; and HCO s - , 12; the osmolality was 293. A 1:50 dilution of this enterotoxin preparation was the minimal amount necessary to cause a positive loop reaction when incubated in the rabbit ileum for 18 hr. In rat cecal inoculations, this toxin was
Downloaded from http://jid.oxfordjournals.org/ at University of Lethbridge on September 6, 2015
Movement of fluid and electrolytes. Movement of cecal fluid and electrolytes was studied in nonfasting Sprague-Dawley rats (250-350 g) by an in vivo closed loop technique [15]. In some of these studies, the Walter Reed variant of the Sprague-Dawley rat of similar weight was used after demonstration that cecal transport of water and electrolytes was similar in the two groups. Briefly, each animal was anesthetized ip with pentobarbital (65 rug/kg body weight). After isolation of the cecum from the ileum and colon by ligatures and washing with isotonic saline, test solutions were inoculated into the cecum, and the cecum was returned to the abdominal cavity. The animals were killed by inhalation of diethyl ether after incubations of 3-18 hr. The cecum was then removed, drained, and washed with isotonic saline to determine the recovery of the nonabsorbable marker (polyethylene glycol). The dry weight of the cecum was determined after heating until a constant weight was achieved (100 C, 18 hr). The initial volume of all solutions placed into the cecum was 3 ml.Both [14C]polyethylene glycol (approximately 0.01 .~Ci/ml) and unlabeled polyethylene glycol (5 g/Iiter ) were added to all solutions. Recovery of polyethylene glycol was 98% ±5% in all groups studied. Net transport of water and electrolytes was calculated as previously described by determination of changes in concentration of polyethylene glycol and expressed as ~l or as ueq/g of dry tissue per min, respectively [15]. Net absorption from the lumen was expressed as a positive value, and net secretion into the lumen as a negative value. Activity of [14C]polyethylene glycol was determined in a liquid scintillation spectrometer. Quench corrections were determined by the method of external standards or channel ratios. Sodium and potassium were measured by flame
Donowitz and Binder
Colonic Effects 0/ Bacterial Enterotoxins
Results
Effects of choleragen on cecum. In preliminary experiments, choleragen (50 I1-g) was inocu-
lated into rat colon, washed out after 2 hr, and studied by a perfusion technique 4 hr after the inoculation of choleragen [24]. Water was secreted in colon inoculated with choleragen (-4.6 -I- 2.9 11-11 em per 20 min) as compared with absorption of water in control colon (8.4 ± 1.2 11-1Icm per 20 min). In all subsequent studies, rat cecum was studied for further definition of the characteristics of the choleragen-induced colonic secretion. When 0.154 M NaCI remained in the rat cecum for 18 hr (table 1), net absorption of water, sodium, and chloride and net secretion of bicarbonate were observed. All doses of choleragen studied produced net secretion of water and sodium, decreased absorption of chloride, and increased secretion of bicarbonate as compared with results in control studies. When the log of the dose of choleragen was plotted against net movement of water and sodium, linear relations between the log of the dose and water movement (r == 0.87; slope == -2.29; P < 0.05) and between the log of the dose and sodium movement (r == 0.84; slope == -0.28; P < 0.05) were found. Heat-inactivated choleragen did not affect movement of water, sodium, chloride, or bicarbonate. When choleragen was incubated in rat cecum for < 18 hr, higher concentrations of choleragen were needed to alter transport of water and electrolytes; concentrations of choleragen of < 50 ug/rnl (table 2) incubated in the rat cecum for 3 hr had no effect on movement of water and electrolytes. For further characterization of the cholerageninduced colonic secretion, the effect of 50 I1-g of choleragerr/ml on cecal transport was studied after inoculation and incubation for 3, 6, 9, and 18 hr (table 3). The colonic response seen 3 hr after inoculation of 50 I1-g of choleragen/rnl was less consistent than that seen after longer periods of incubation. Net secretion of water and sodium occurred in only nine of 12 experiments in which choleragen was incubated for 3 hr. During 6-, 9-, and 18-hr incubations, net secretion of water was always observed (table 3). When choleragen was incubated for 3-18 hr in rat ceca, the effect of choleragen on net movement of water, sodium, chloride, and bicarbonate, as calculated from the difference between choleragen- and saline-inoculated ceca, was similar at all times. For example, the difference in net movement of water between
Downloaded from http://jid.oxfordjournals.org/ at University of Lethbridge on September 6, 2015
used either undiluted or diluted 1: 10 in Ringer's bicarbonate solution. Lyophilized S. dysenteriae type 1 enterotoxin was produced by a previously described method [19] and kindly supplied by Dr. G. T. Keusch (Mt, Sinai School of Medicine, New York, N.Y.). The shigella enterotoxin (Keusch lot no. Sa-B) was standardized with the HeLa cell model; the minimal cidal dose for HeLa cells was 0.1 ug/rnl [20]. A cecal inoculum consisted of 1 mg of enterotoxin in Ringer's bicarbonate solution. Control solutions for the E. coli enterotoxins and S. dysenteriae type 1 enterotoxin was Ringer's bicarbonate solution, with a composition in meq/Iiter, as follows: Na, 140; K, 5.2; Mg, 1.2; Ca, 1.2; CI, 119.8; HC0 3 - , 25; HP0 4 - 2 , 2.4; and H 2P04 - , 0.4. Mucosal cyclic AMP content. The cyclic AMP content of the cecal mucosa was determined in a separate series of in vivo experiments. N aCI (0.154 M; 3 ml) or choleragen (50 ug/rnl) was placed into the cecum. After incubation of the cecum for 3 hr, the animal was anesthetized with ether, and the cecum was removed while the animal was still alive. The mucosa was scraped with a glass slide, and the tissue was frozen immediately in liquid nitrogen; no longer than 60 sec elapsed between induction of anesthesia and freezing of the tissue. The concentration of cyclic AMP was measured by the protein-binding method of Brown et al. [21]. The protein concentration in the precipitate was determined by the method of Lowry et al. [22]. Cyclic AMP was expressed as pg/mg of protein. Histology. After incubation with choleragen, E. coli enterotoxin, or S. dysenteriae type 1 enterotoxin, ceca were removed for histologic sections. Formalin-fixed sections stained with hematoxylin and eosin, periodic acid-Schiff reagent, and alcian blue were coded and examined. The experimental results were analyzed for statistical significance by Student's t-test, and linear regression analysis was performed by the method of least squares [23]. All values represent the mean -I- SE.
137
Donowitz and Binder
138
Table 1. Effect of choleragen dose on net cecal water and electrolyte movement after incubation for 18 hr in ceca of indicated number of rats. Movement of Dose of choleragen (ug /rnl)
No. of rats Water (f,tUg of tested dry weight per min)
o
14 5 6 4 6 5 5 6 5
5.9 ± 1.8 6.1 ± 2.1 -10.9 ± 3.3* -2.2 ± 3.0:1: -10.9 ± 1.9§ -5.0 ± 2.3t -7.8 ± 2.3§ -21.7 ± 3.3§ -24.8±4.1§
1.02 ± 2.60 ± -0.98 ± 0.01 ± -1.11 ± -0.13 ± -0.48 ± -2.24 ± -2.84 ±
0.10 0.80 0.51 t 0.63 0.24§ 0.19§ 0.31 § 0.35§ 0.53§
Chloride (ueq Zgof dry weight per min) 1.84 ± 0.16 1.42 ± 0.43 1.09 ± 0.15* 1.23 ± 0.14:1: 0.95 ± 0.10* 1.28 ± 0.11§ 1.05 ± 0.17§ 0.62 ± 0.07§ 0.33±0.11§
Bicarbonate (f,teq/g of dry weight per min) -0.97 ± -1.10 ± -2.26 ± -1.84 ± -2.41 ± -1.85 ± -2.33 ± -4.22 ± -4.03 ±
0.07 0.43 0.35t 0.43 0.44* 0.29:1: 0.45:1: 0.22§ 0.47§
NOTE. Results are expressed as mean ± SE. Positive values represent net absorption from lumen, and negative values net secretion into lumen. P values represent comparisons between saline- and choleragen-inoculated ceca. * P < 0.01. t F < 0.005. :I: P < 0.05. § P < 0.001.
choleragen- and saline-inoculated ceca at 3 hr (24.5 ± 5.7 fll/g of dry weight per min) was not significantly different from the difference in net movement of water at 18 hr (27.6 ± 5.0 ul/g of dry weight per min). Levels of cyclic AMP were also measured in cecal mucosa exposed to choleragen (50 flg/ ml) for 3 hr and were significantly greater in the experimental animals (11.22 ± 0.94 pmol/rng of protein) than in control tissue (5.95 ± 0.33 pmols/rng of protein; P < 0.01). Histologic sections of ceca incubated with saline solution or choleragen (50 ug/rnl) for 3, 6, and 18 hr were obtained from parallel experiments, coded, and evaluated independently by two observers. In the tissue exposed to choleragen for 18 hr, the surface epithelial cells were flattened as compared with those in control ~ tissues. No otherhistologic alterations were present in chol-
eragen-exposed tissues. Choleragen did not affect the amount of colonic goblet-cell mucus as judged after staining with periodic acid-Schiff reagent and aldan blue. Cecal effects of E. coli enterotoxin. The effects of two different preparations of E. coli enterotoxin were studied to determine whether there was an enterotoxin-mediated alteration of colonic transport of water and electrolytes (table 4). An enterotoxin prepared from a porcine enteropathogenic strain of E. coli that contained only the highmolecular-weight form of E. coli enterotoxin did not affect cecal transport of water and electrolytes when a dose twice that needed to produce fluid secretion in rabbit ileal loops was incubated for 3, 6, and 18 hr. Similarly, a sonicate of E. coli enterotoxin that contained both the heat-labile and heat-stable forms of enterotoxin also failed to alter cecal water and electrolyte transport when
Table 2. Effect of choleragen dose on net cecal water and sodium movement after incubation for 3 hr in ceca of indicated number of rats. Dose of choleragen (ug /rnl )
o 16.7 50
No. of rats tested
Movement of water (ul /g of dry weight per min)
14 6 12
11.7±2.6 18.4 ± 2.7 -12.8 ± 5.1
P
Movement of sodium (ueqz'g of dry weight per min)
P
NS
AUGUST 1976
Effect of Enterotoxins of Vibrio cholerae, Escherichia coli, and Shigella dysenteriae Type 1 on Fluid and Electrolyte Transport in the Colon From the Department of Gastroenterology, Walter Reed Army In;titute of Research, Washington, D.C.; and the Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut
Mark Donowltz and Henry J. Binder
Recent studies of the pathogenesis of bacterial diarrhea have identified intestinal secretion of electrolytes in diseases associated with entero-
toxins as well as in diseases for. which no such association has been found [1, 2]. At least seven bacterial enterotoxins have been identified, and, by definition, all induce secretion of fluid by the small intestine [1]. In both cholera and diarrhea associated with noninvasive strains of Escherichia coli, the respective enterotoxins cause secretion of fluid by the small intestine and stimulate adenylate cyclase both in the small intestine and in nonintestinal tissue [1, 2]. Fluid and electrolytes are also secreted during non-enterotoxinassociated bacterial diarrheas such as salmonellosis [3]. In shigellosis, although an enterotoxin has been identified, it is uncertain whether this enterotoxin is involved in the production of diarrhea. In contrast to the extensive evaluation of the effects of the bacterial enterotoxins on the small intestine, it has not been determined whether the colon is involved in the enterotoxin-mediated bacterial diarrheas. Several communications have suggested that Vibrio cholerae enterotoxin does not affect movement of fluid and electrolytes in the colon [4-6]. These observations are surprising in light of the known ability of cholera toxin to increase levels of adenyl ate cyclase in many tissues other than the small intestine (i.e., thyroid, leukocytes, and erythrocytes) [7-10]. In addition, recent observations have shown that adenosine
Received for publication October 16, 1975, and in revised form February 9, 1976. This work was supported by grant no. AM 14669 from the National Institute of Arthritis, Metabolism, and Digestive Diseases, and by grants from the John A. Hartford Foundation, Inc., and the Connecticut Digestive Disease Society. Dr. Donowitz was supported by training grant no. AM 05703 from the National Institute of Arthritis, Metabolism, and Digestive Diseases. We acknowledge the expert technical assistance of J. M. Heffernan, P. Fisher, D. Whiting, and G.Chapo, and the secretarial assistance of Carol Bryan and Elizabeth Penn. We thank Drs. Carl E. Miller (National Institute of Allergy and Infectious Diseases), Thomas M. Jacks (Norwich Pharmacal Company, Norwich, N. Y.), Samuel Formal (Walter Reed Army Institute of Research), and Gerald T. Keusch (Mt. Sinai School of Medicine, New York, N.Y.) for their kind gifts of enterotoxins. In conducting the research described in this paper, the investigators adhered to the "Guide for Laboratory Animal Facilities and Care," as promulgated by the Committee on the Guide for Laboratory Animal Facilities and Care of the Institute of Laboratory Animal Resources, National Academy of Sciences-National Research Council. Please address request for reprints to Dr. Mark Donowitz, Department of Gastroenterology, Walter Reed Army Institute of Research, Washington, D.C. 20012.
135
Downloaded from http://jid.oxfordjournals.org/ at University of Lethbridge on September 6, 2015
Experiments were designed to determine whether the enterotoxins of Vibrio cholerae, Escherichia coli, and Shigella dysenteriae type 1. alter the movement of fluid and electrolytes in the rat cecum. Net secretion of water and sodium were observed after incubation of 1.67 X 10- 5 ltg of purified cholera toxin (choleragen)/ml for 18 hr or of 50 ltg/ ml for 3 hr. The effect of choleragen on cecal transport of water and electrolytes was related to the dose. In addition, choleragen increased cecal mucosal content of adenosine 3': 5'-cyclic phosphate but did not alter the histology of the cecum. The results demonstrate that the colon responds to choleragen in a manner similar to that. of. other tissues. In contrast, the enterotoxins of both E. coli and S. dysenteriae type 1 failed to affect cecal transport of water and electrolytes. These observations may explain several phenomena associated with the diarrhea produced by bacterial enterotoxins.
136
3' :5'-cyclic phosphate (cyclic AMP) induces the secretion of electrolytes in the colon [11, 12], and that guinea pig and rabbit colons bind large amounts of choleragen [13, 14]. The present studies were designed to evaluate whether the enterotoxins of V. cholerae as well as those of E. coli and Shigella dysenteriae type 1 affect transport of fluid and electrolytes by the colon. Materials and Methods
photometry (model no, 143, Instrumentation Laboratories, Woburn, Mass.), and chloride by coulometric titration (Buchler Instrument Division, Fort Lee, N.J.). The concentration of bicarbonate was calculated by measurement of the difference between the concentration of sodium plus potassium and that of chloride. Osmolality was determined by freezing point depression with an osmometer (Advanced Instruments, Needham Heights, Mass.). Enterotoxin preparations. The effect of cholera enterotoxin on the colon was evaluated by inoculation into the cecum of purified cholera enterotoxin (choleragen; lot no. 0972) prepared by Dr. Richard A. Finkelstein (University of Texas Southwestern Medical School, Dallas, Tx.) [16] and generously provided by Dr. Carl E. Miller (National Institutes of Health). The concentration of choleragen, made up in 0.154 M NaCl, ranged from 1.67 X 10- 5 ug/rnl to 166.7 ug/rnl, Heatinactivated (l00 C, 30 min) choleragen (50 ug/rnl) was also studied. The control solution was 0.154 M NaCl. Two different E. coli exterotoxins were studied. (1) A lyophilized, crude, high-molecular-weight enterotoxin prepared from a porcine enteropathogenic strain of E. coli (08 :K87, 88a,b:HI9) [17] was kindly supplied by Dr. T. M. Jacks (Norwich Pharmacal Co., Norwich, N.Y.). This enterotoxin had been shown to contain only the high-molecular-weight, heat-labile form of E. coli enterotoxin; 7.3 mg was the minimal amount necessary to cause a positive loop reaction when incubated for 18 hr in the rabbit ileum. Twice this amount of toxin (5 mg/rnl) , made up in Ringer's bicarbonate solution, was inoculated into rat cecum. (2) Enterotoxin prepared from a porcine enteropathogenic strain of E. coli (08:K87, K88:HI9) was kindly supplied as a whole-cell lysate by Dr. Samuel Formal (Walter Reed Army Institute of Research). This enterotoxin had been shown previously to contain both the heat-labile and heatstable forms of E. coli enterotoxin [18]. The electrolyte concentration of this material, in meq/liter, was: Na, 156; K, 2.8; CI, 146.8; and HCO s - , 12; the osmolality was 293. A 1:50 dilution of this enterotoxin preparation was the minimal amount necessary to cause a positive loop reaction when incubated in the rabbit ileum for 18 hr. In rat cecal inoculations, this toxin was
Downloaded from http://jid.oxfordjournals.org/ at University of Lethbridge on September 6, 2015
Movement of fluid and electrolytes. Movement of cecal fluid and electrolytes was studied in nonfasting Sprague-Dawley rats (250-350 g) by an in vivo closed loop technique [15]. In some of these studies, the Walter Reed variant of the Sprague-Dawley rat of similar weight was used after demonstration that cecal transport of water and electrolytes was similar in the two groups. Briefly, each animal was anesthetized ip with pentobarbital (65 rug/kg body weight). After isolation of the cecum from the ileum and colon by ligatures and washing with isotonic saline, test solutions were inoculated into the cecum, and the cecum was returned to the abdominal cavity. The animals were killed by inhalation of diethyl ether after incubations of 3-18 hr. The cecum was then removed, drained, and washed with isotonic saline to determine the recovery of the nonabsorbable marker (polyethylene glycol). The dry weight of the cecum was determined after heating until a constant weight was achieved (100 C, 18 hr). The initial volume of all solutions placed into the cecum was 3 ml.Both [14C]polyethylene glycol (approximately 0.01 .~Ci/ml) and unlabeled polyethylene glycol (5 g/Iiter ) were added to all solutions. Recovery of polyethylene glycol was 98% ±5% in all groups studied. Net transport of water and electrolytes was calculated as previously described by determination of changes in concentration of polyethylene glycol and expressed as ~l or as ueq/g of dry tissue per min, respectively [15]. Net absorption from the lumen was expressed as a positive value, and net secretion into the lumen as a negative value. Activity of [14C]polyethylene glycol was determined in a liquid scintillation spectrometer. Quench corrections were determined by the method of external standards or channel ratios. Sodium and potassium were measured by flame
Donowitz and Binder
Colonic Effects 0/ Bacterial Enterotoxins
Results
Effects of choleragen on cecum. In preliminary experiments, choleragen (50 I1-g) was inocu-
lated into rat colon, washed out after 2 hr, and studied by a perfusion technique 4 hr after the inoculation of choleragen [24]. Water was secreted in colon inoculated with choleragen (-4.6 -I- 2.9 11-11 em per 20 min) as compared with absorption of water in control colon (8.4 ± 1.2 11-1Icm per 20 min). In all subsequent studies, rat cecum was studied for further definition of the characteristics of the choleragen-induced colonic secretion. When 0.154 M NaCI remained in the rat cecum for 18 hr (table 1), net absorption of water, sodium, and chloride and net secretion of bicarbonate were observed. All doses of choleragen studied produced net secretion of water and sodium, decreased absorption of chloride, and increased secretion of bicarbonate as compared with results in control studies. When the log of the dose of choleragen was plotted against net movement of water and sodium, linear relations between the log of the dose and water movement (r == 0.87; slope == -2.29; P < 0.05) and between the log of the dose and sodium movement (r == 0.84; slope == -0.28; P < 0.05) were found. Heat-inactivated choleragen did not affect movement of water, sodium, chloride, or bicarbonate. When choleragen was incubated in rat cecum for < 18 hr, higher concentrations of choleragen were needed to alter transport of water and electrolytes; concentrations of choleragen of < 50 ug/rnl (table 2) incubated in the rat cecum for 3 hr had no effect on movement of water and electrolytes. For further characterization of the cholerageninduced colonic secretion, the effect of 50 I1-g of choleragerr/ml on cecal transport was studied after inoculation and incubation for 3, 6, 9, and 18 hr (table 3). The colonic response seen 3 hr after inoculation of 50 I1-g of choleragen/rnl was less consistent than that seen after longer periods of incubation. Net secretion of water and sodium occurred in only nine of 12 experiments in which choleragen was incubated for 3 hr. During 6-, 9-, and 18-hr incubations, net secretion of water was always observed (table 3). When choleragen was incubated for 3-18 hr in rat ceca, the effect of choleragen on net movement of water, sodium, chloride, and bicarbonate, as calculated from the difference between choleragen- and saline-inoculated ceca, was similar at all times. For example, the difference in net movement of water between
Downloaded from http://jid.oxfordjournals.org/ at University of Lethbridge on September 6, 2015
used either undiluted or diluted 1: 10 in Ringer's bicarbonate solution. Lyophilized S. dysenteriae type 1 enterotoxin was produced by a previously described method [19] and kindly supplied by Dr. G. T. Keusch (Mt, Sinai School of Medicine, New York, N.Y.). The shigella enterotoxin (Keusch lot no. Sa-B) was standardized with the HeLa cell model; the minimal cidal dose for HeLa cells was 0.1 ug/rnl [20]. A cecal inoculum consisted of 1 mg of enterotoxin in Ringer's bicarbonate solution. Control solutions for the E. coli enterotoxins and S. dysenteriae type 1 enterotoxin was Ringer's bicarbonate solution, with a composition in meq/Iiter, as follows: Na, 140; K, 5.2; Mg, 1.2; Ca, 1.2; CI, 119.8; HC0 3 - , 25; HP0 4 - 2 , 2.4; and H 2P04 - , 0.4. Mucosal cyclic AMP content. The cyclic AMP content of the cecal mucosa was determined in a separate series of in vivo experiments. N aCI (0.154 M; 3 ml) or choleragen (50 ug/rnl) was placed into the cecum. After incubation of the cecum for 3 hr, the animal was anesthetized with ether, and the cecum was removed while the animal was still alive. The mucosa was scraped with a glass slide, and the tissue was frozen immediately in liquid nitrogen; no longer than 60 sec elapsed between induction of anesthesia and freezing of the tissue. The concentration of cyclic AMP was measured by the protein-binding method of Brown et al. [21]. The protein concentration in the precipitate was determined by the method of Lowry et al. [22]. Cyclic AMP was expressed as pg/mg of protein. Histology. After incubation with choleragen, E. coli enterotoxin, or S. dysenteriae type 1 enterotoxin, ceca were removed for histologic sections. Formalin-fixed sections stained with hematoxylin and eosin, periodic acid-Schiff reagent, and alcian blue were coded and examined. The experimental results were analyzed for statistical significance by Student's t-test, and linear regression analysis was performed by the method of least squares [23]. All values represent the mean -I- SE.
137
Donowitz and Binder
138
Table 1. Effect of choleragen dose on net cecal water and electrolyte movement after incubation for 18 hr in ceca of indicated number of rats. Movement of Dose of choleragen (ug /rnl)
No. of rats Water (f,tUg of tested dry weight per min)
o
14 5 6 4 6 5 5 6 5
5.9 ± 1.8 6.1 ± 2.1 -10.9 ± 3.3* -2.2 ± 3.0:1: -10.9 ± 1.9§ -5.0 ± 2.3t -7.8 ± 2.3§ -21.7 ± 3.3§ -24.8±4.1§
1.02 ± 2.60 ± -0.98 ± 0.01 ± -1.11 ± -0.13 ± -0.48 ± -2.24 ± -2.84 ±
0.10 0.80 0.51 t 0.63 0.24§ 0.19§ 0.31 § 0.35§ 0.53§
Chloride (ueq Zgof dry weight per min) 1.84 ± 0.16 1.42 ± 0.43 1.09 ± 0.15* 1.23 ± 0.14:1: 0.95 ± 0.10* 1.28 ± 0.11§ 1.05 ± 0.17§ 0.62 ± 0.07§ 0.33±0.11§
Bicarbonate (f,teq/g of dry weight per min) -0.97 ± -1.10 ± -2.26 ± -1.84 ± -2.41 ± -1.85 ± -2.33 ± -4.22 ± -4.03 ±
0.07 0.43 0.35t 0.43 0.44* 0.29:1: 0.45:1: 0.22§ 0.47§
NOTE. Results are expressed as mean ± SE. Positive values represent net absorption from lumen, and negative values net secretion into lumen. P values represent comparisons between saline- and choleragen-inoculated ceca. * P < 0.01. t F < 0.005. :I: P < 0.05. § P < 0.001.
choleragen- and saline-inoculated ceca at 3 hr (24.5 ± 5.7 fll/g of dry weight per min) was not significantly different from the difference in net movement of water at 18 hr (27.6 ± 5.0 ul/g of dry weight per min). Levels of cyclic AMP were also measured in cecal mucosa exposed to choleragen (50 flg/ ml) for 3 hr and were significantly greater in the experimental animals (11.22 ± 0.94 pmol/rng of protein) than in control tissue (5.95 ± 0.33 pmols/rng of protein; P < 0.01). Histologic sections of ceca incubated with saline solution or choleragen (50 ug/rnl) for 3, 6, and 18 hr were obtained from parallel experiments, coded, and evaluated independently by two observers. In the tissue exposed to choleragen for 18 hr, the surface epithelial cells were flattened as compared with those in control ~ tissues. No otherhistologic alterations were present in chol-
eragen-exposed tissues. Choleragen did not affect the amount of colonic goblet-cell mucus as judged after staining with periodic acid-Schiff reagent and aldan blue. Cecal effects of E. coli enterotoxin. The effects of two different preparations of E. coli enterotoxin were studied to determine whether there was an enterotoxin-mediated alteration of colonic transport of water and electrolytes (table 4). An enterotoxin prepared from a porcine enteropathogenic strain of E. coli that contained only the highmolecular-weight form of E. coli enterotoxin did not affect cecal transport of water and electrolytes when a dose twice that needed to produce fluid secretion in rabbit ileal loops was incubated for 3, 6, and 18 hr. Similarly, a sonicate of E. coli enterotoxin that contained both the heat-labile and heat-stable forms of enterotoxin also failed to alter cecal water and electrolyte transport when
Table 2. Effect of choleragen dose on net cecal water and sodium movement after incubation for 3 hr in ceca of indicated number of rats. Dose of choleragen (ug /rnl )
o 16.7 50
No. of rats tested
Movement of water (ul /g of dry weight per min)
14 6 12
11.7±2.6 18.4 ± 2.7 -12.8 ± 5.1
P
Movement of sodium (ueqz'g of dry weight per min)
P
NS
