Effect of Bacterial Lipopolysaccharide on Gastric Emptying of Liquids in Rats

Home , Lipopolysaccharide

BrazilianEndotoxin Journal and gastric of Medical emptying and Biological Research (1997) 30: 207-211 207 ISSN 0100-879X Short Communication

Effect of bacterial lipopolysaccharide on gastric emptying of liquids in rats

E.F. Collares Departamento de Pediatria and Núcleo de Medicina e Cirurgia Experimental, Faculdade de Ciências Médicas, Universidade Estadual de Campinas, 13081-970 Campinas, SP, Brasil

Abstract

Correspondence The objectives of the present investigation were 1) to study the effect Key words E.F. Collares of bacterial lipopolysaccharide (LPS) on rat gastric emptying (GE) and • Gastric emptying Departamento de Pediatria 2) to investigate a possible involvement of the vagus nerve in the • Lipopolysaccharide Faculdade de Ciências Médicas gastric action of LPS. Endotoxin from E. coli (strain 055:B5) was • Vagus nerve Universidade Estadual de Campinas administered sc, ip or iv to male Wistar rats (220-280 g body weight) 13081-970 Campinas, SP Brasil at a maximum dose of 50 µg/kg animal weight. Control animals received an equivalent volume of sterile saline solution. At a given Publication supported by FAPESP. time period after LPS administration, GE was evaluated by measuring gastric retention 10 min after the orogastric infusion of a test meal (2 ml/100 g animal weight), which consisted of 0.9% NaCl plus the marker phenol red (6 mg/dl). One group of animals was subjected to Received March 12, 1996 bilateral subdiaphragmatic vagotomy or sham operation 15 days be- Accepted December 5, 1996 fore the test. A significant delay in GE of the test meal was observed 5 h after iv administration of the endotoxin at the dose of 50 µg/kg animal weight. The LPS-induced delay of GE was detected as early as 30 min and up to 8 h after endotoxin administration. The use of different doses of LPS ranging from 5 to 50 µg/kg animal weight showed that the alteration of GE was dose dependent. In addition, vagotomized animals receiving LPS displayed a GE that was not significantly different from that of the sham control group. However, a participation of the vagus nerve in LPS-induced delay in GE could not be clearly demonstrated by these experiments since vagotomy itself induced changes in this gastric parameter. The present study provides a suitable model for identifying the mechanisms underlying the effects of LPS on gastric emptying.

Natural infection or parenterally admin- diarrhea (4). In addition, LPS at sublethal istered lipopolysaccharide (LPS) of bacterial doses suppresses gastric motility as suggested origin can result in various clinical manifes- by the LPS-induced inhibition of gastric emp- tations in several animal species (1-3). In the tying of liquids and solids observed in mice gastrointestinal tract, bacterial endotoxins and rats (7,8). The mechanisms underlying induce severe intestinal mucosal injury fre- the toxic actions of endotoxins on gastric quently accompanied by diarrhea and enteric emptying remain unexplored. bacterial translocation (4-6). It has been sug- The aim of the present study was to ex- gested that vascular damage may play a role amine the effect of LPS from E. coli on in the pathogenesis of LPS-induced acute gastric emptying of a saline solution in rats

Braz J Med Biol Res 30(2) 1997 208 E.F. Collares

evaluating the optimum route of administra- ume of 2 ml/100 g animal weight. The test tion, dosage and time following administra- meal consisted of an aqueous solution of tion of the endotoxin. After this preliminary 0.9% NaCl containing 6 mg/dl phenol red as phase of the study, the second part was set up a marker. Saline solution has been consid- to determine vagal involvement in the ered a suitable test meal to evaluate gastric LPS-induced alteration of gastric emptying. emptying without intestinal interference since Male Wistar rats (SPF, Biotério Central, there is no activation of intestinal receptors Universidade Estadual de Campinas, Brazil) in contrast to that obtained with nutritive test aged 8-10 weeks and weighing 220-280 g meals (11). were employed. The rats were allowed to In addition, a group of animals were sub- habituate to the laboratory environment for jected to bilateral subdiaphragmatic va- at least 15 days before any experimental gotomy 15 days before the test day (12). manipulation. The animals were collectively Sham surgery was also perfomed in some housed and kept under conditions of con- animals by executing laparotomy without trolled temperature (22-26°C) and artificial the section of the vagus (12). Data for ani- dark-light cycle of 12 h. They were fed on a mals displaying severe gastric stasis after standard rat chow (Labina, Purina) and re- vagotomy were excluded from data analysis. ceived water ad libitum. Data are reported as means ± SEM (num- Before the test day, the rats were trans- ber of animals). The Student t-test (k = 2) or ferred to individual cages and subjected to a analysis of variance (k ≥ 3) followed by the 20-h fasting period during which only water Tukey test were used for statistical analysis was available. LPS from E. coli (strain of the data, with the level of significance set 055:B5, Sigma) was dissolved in 0.9% at P<0.05. Correlation coefficients were cal- nonpyrogenic sterile NaCl and administered culated for the dose-response data. at a maximum dose of 50 µg/kg animal The results of the present study showed weight. Three different administration routes that 5 h after iv administration of LPS (50 µg/ were employed for LPS injection: intrave- kg animal weight), the treated animals pre- nous (iv) using the tail vein, subcutaneous sented a significantly higher percentage of (sc) at the external region of the posterior GR of the test meal than the control group limbs and intraperitoneal (ip). Control injec- (%GR after 10 min, control = 34.8 ± 3.2% (N tions consisted of an equivalent volume of = 8), treated = 63.6 ± 2.3% (N = 8); P<0.05, the appropriate vehicle (1 ml/kg animal t-test). In contrast, when the endotoxin was weight). administered sc or ip using the same experi- Gastric emptying was evaluated 5 h after mental protocol as employed above, no sta- parenteral administration of the endotoxin, tistical difference in GR was observed be- except in the experiment testing the time tween the control and treated groups (%GR dependence of the gastric response to LPS after 10 min, control sc = 33.2 ± 2.1% (N = when the measurement of gastric emptying 8) vs treated sc = 33.7 ± 2.2% (N = 8); was performed at different times (up to 24 h) control ip = 33.7 ± 2.2% (N = 8) vs treated ip after iv injection of endotoxin. Gastric emp- = 41.1 ± 2.2% (N = 8)). tying was indirectly evaluated by determin- To determine the dose dependence of ing the percentage of gastric retention (%GR) this LPS-induced alteration of gastric emp- of a phenol red-containing test meal recov- tying, different doses of LPS (ranging from 5 ered within the gastric content as compared to 50 µg/kg animal weight) injected only iv to that orally administered (9,10). Gastric were tested. As shown in Figure 1A, there retention was measured 10 min after was a close correlation between increasing orogastric infusion of the test meal in a vol- doses of LPS and the intensity of the in-

Braz J Med Biol Res 30(2) 1997 Endotoxin and gastric emptying 209 crease of gastric retention induced by the Figure 1 - Effect of E. coli lipo- 80 endotoxin (correlation coefficient, r = 0.97). polysaccharide on gastric reten- A tion of 0.9% NaCl by rats. A, * Among the doses tested, 25 and 50 µg/kg of Gastric retention was evaluated 60 LPS were effective in significantly increas- 5 h after the iv administration of * ing GR of a saline solution in rats (P<0.05 different LPS doses. B, Gastric retention was determined 0.5, 40 compared to the control group, Tukey test). 2, 5, 8 and 24 h after iv adminis- Figure 1B shows the time-response ef- tration of 50 µg/kg LPS. Control groups received vehicle instead fect of the endotoxin. For this set of experi- Gastric retention (%) 20 of endotoxin. Data are reported ments, GR was determined 0.5, 2, 5, 8 and 24 as means ± SEM for 8 animals h after iv administration of 50 µg/kg LPS. 0 (panel A) or 6 animals (panel B) The endotoxin induced a rapid and sustained 0 5 12.5 25 50 in each subgroup. *P<0.05 com- LPS doses (µg/kg) pared to control group (Tukey increase in GR of the saline solution since 100 test). r = 0.97 for dose-response significant differences between the treated Control correlation. +P<0.05 compared LPS-treated B to control group (Tukey test). group and its respective controls (P<0.05, 80 ++++ Tukey test) were observed as early as 30 min and up to 8 h after endotoxin administration. 60 However, LPS was not effective in inducing 40 this gastric effect 24 h after its administration (Figure 1B). Gastric retention (%) 20 In the last part of the present study, we 0 determined whether vagal innervation of the 0.5 2 5 8 24 stomach was involved in the LPS-induced Time after LPS injection (h) delay of gastric emptying. Figure 2 shows the results obtained. When the two subgroups Figure 2 - Effect of vagotomy on treated with the endotoxin for 5 h were com- 100 Control the E. coli lipopolysaccharide pared, the vagotomized animals showed sig- LPS-treated (LPS)-induced increase in gastric * nificantly lower %GR values than the 80 retention of 0.9% NaCl solution sham-operated animals (P<0.05, Tukey test). * by rats. The animals were sub- 60 jected to bilateral subdiaphrag- These data suggest that the vagus might me- matic vagotomy 15 days before diate, at least partially, the action of LPS 40 the test. Gastric retention was upon gastric emptying. However, vagotomy evaluated 5 h after the iv admin- Gastric retention (%) 20 istration of LPS (50 µg/kg animal itself had an effect on GR as indicated by the weight). The control subgroups fact that vagotomized control animals dis- received vehicle instead of en- 0 played significantly lower %GR values than Sham Vagotomy dotoxin. Data are reported as the sham control group (P<0.05, Tukey test). means ± SEM for 8 animals in each subgroup. The %GR val- In conclusion, the present data clearly “normalize” the gastric emptying process in ues between control and treated demonstrate that LPS from E. coli induces a rats treated with LPS since their %GR values groups were significantly differ- significant delay of gastric emptying of an were partially restored to control values (Fig- ent. *P<0.05 compared to the indicated experimental group aqueous solution in rats. This alteration of ure 2 compares sham control vs sham-treated (Tukey test). gastric function was closely dependent on and sham control vs vagotomized-treated the administration route, dosage and time groups). These data suggest that the vagus after administration of the endotoxin. This nerve may play a role in the endotoxin -in- gastric disturbance was observed with con- duced change in gastric emptying. It is well siderably lower doses of LPS as compared to established that afferent and efferent vagal those employed in previous studies (7,8). innervation of the stomach modulates the Concerning the vagal involvement in the gastric motility and function via vagal cho- phenomenon, bilateral vagotomy seems to linergic excitatory fibers and non-adrenergic

Braz J Med Biol Res 30(2) 1997 210 E.F. Collares

and non-cholinergic inhibitory fibers (13). bacterial endotoxin on specific areas of the The faster gastric emptying of liquids ob- CNS which control some of the gastric func- served in the present study after sectioning tions via the vagus nerve. Several studies of the vagus trunk has also been described have demonstrated that certain hypothalam- by others (13). However, the above conclu- ic nuclei, e.g., the paraventricular (12) and sion drawn from the present data seems not dorsomedial nuclei (15), can modulate the to be straightforward since vagotomy itself motor activity of the gastrointestinal tract. In had an effect on gastric emptying. In this addition, several reports have provided evi- respect, one may suggest that the sectioning dence for a vagal and central action of LPS. of the vagus nerve does not block the effect For instance, endotoxin from E. coli can of endotoxin on gastric emptying since the induce 1) a fever response in rabbits when injection of LPS induced a two-fold increase administered intracerebroventricularly (16), in %GR in the vagotomized group (vagoto- and 2) activation of hypothalamic histamin- mized control %GR = 19.1%, vagotomized ergic neurons in rats leading to ACTH re- LPS = 38.4%), similar to the values ob- lease (17). In addition, a recent study has served in the sham group (sham control demonstrated that subdiaphragmatic va- %GR = 34.2%, sham LPS = 55.6%). gotomy can suppress the endotoxin-induced Further work is needed to identify the activation of hypothalamic corticotropin -re- exact mechanisms by which the E. coli en- leasing hormone and ACTH secretion (18). dotoxin affects gastric motility. An attrac- tive possibility is that LPS induces alter- Acknowledgments ations in the paracrine control of gastric emptying via locally released cytokines. For The author would like to thank Dr. Carla instance, endotoxin-induced intestinal vas- B. Collares-Buzato (Departamento de cular damage is known to be mediated by Fisiologia e Biofísica, UNICAMP, Brazil) cytokines such as nitric oxide, platelet- acti- for critically reviewing the English text, Ms. vating factor and thromboxane A2 (14). Al- Maria Cristina Fernandes Alvin for techni- ternatively, the LPS-induced gastric distur- cal support and Ms. Irani Ribeiro for typing bance may be the result of an action of this the manuscript.

Braz J Med Biol Res 30(2) 1997 Endotoxin and gastric emptying 211

References

1. Stetson Jr CA (1961). Symposium on bac- 9. Belangero VMS & Collares EF (1991). 15. Greenwood B & DiMicco JA (1995). Acti- terial endotoxins. IV. Immunological as- Esvaziamento gástrico e acidose metabó- vation of the hypothalamic dorsomedial pects of the host reaction to endotoxins. lica. I. Estudo de um modelo experimen- nucleus stimulates intestinal motility in Bacteriological Reviews, 25: 457-458. tal em ratos, empregando uma solução rats. American Journal of Physiology, 268 2. Zweifach BW & Janoff A (1965). Bacterial de cloreto de amônio por via orogástrica. (Gastrointestinal and Liver Physiology, endotoxemia. Annual Review of Medi- Arquivos de Gastroenterologia, 28: 145- 31): G514-G521. cine, 16: 201-220. 150. 16. Macari M, Pela IR, Silva CAA & Viana RS 3. Kluger MJ (1991). Fever: Role of pyrogens 10. Bucaretchi F & Collares EF (1996). Effect (1990). Fever response induced by intra- and cryogens. Physiological Reviews, 71: of Phoneutria nigriventer spider venom venous and intracerebroventricular injec- 93-127. on gastric emptying in rats. Brazilian Jour- tion of pyrogen in thyroidectomised and 4. Mathan VI, Penny GR, Mathan MM & nal of Medical and Biological Research, protein-calorie malnourished rabbits. Rowlewy D (1988). Bacterial lipopolysac- 29: 205-211. Pflügers Archiv: European Journal of charide-induced intestinal microvascular 11. Hunt JN (1983). Mechanisms and disor- Physiology, 15: 440-443. lesions leading to acute diarrhea. Journal ders of gastric emptying. Annual Reviews 17. Knigge U, Kjaer A, Jorgensen H, Garbarg of Clinical Investigation, 82: 1714-1721. of Medicine, 31: 219-229. M, Ross C, Rouleau A & Warberg J (1994). 5. Fink MP (1991). Gastrointestinal mucosal 12. Collares-Buzato CB, Collares EF & Role of hypothalamic histaminergic neu- injury in experimental models of shock, Fernandes GA (1993). Effect of paraven- rons in mediation of ACTH and beta-en- trauma, and sepsis. Critical Care Medi- tricular nucleus lesion and cold restraint dorphin responses to LPS endotoxin in cine, 19: 627-641. stress on gastric emptying of a liquid meal vivo. Neuroendocrinology, 60: 243-251. 6. Wells CL, Barton RG, Wavatne CS, Dunn in rats. Brazilian Journal of Medical and 18. Gaykema RPA, Dijkstra I & Tilders FJH DL & Cerra FB (1992). Intestinal bacterial Biological Research, 26: 1009-1014. (1995). Subdiaphragmatic vagotomy sup- flora, intestinal pathology, and lipopoly- 13. Kelly KA (1984). Effect of gastric surgery presses endotoxin-induced activation of saccharide-induced translocation of intes- on gastric motility and emptying. In: hypothalamic corticotropin-releasing hor- tinal bacteria. Circulatory Shock, 37: 117- Akkermans LMA, Johnson AG & Read mone neurons and ACTH secretion. En- 123. NW (Editors), Gastric and Gastroduodenal docrinology, 136: 4717-4720. 7. Turner MM & Berry LJ (1963). Inhibition Motility. Praeger Publishers, New York. of gastric emptying in mice by bacterial 14. Stark ME & Szurszewshi JH (1992). Role endotoxins. American Journal of Physiol- of nitric oxide in gastrointestinal and he- ogy, 205: 1113-1116. patic function and disease. Gastroenterol- 8. Langhans W, Harlacher R, Balkowski G & ogy, 103: 1928-1949. Scharrer E (1990). Comparison of the effects of bacterial lipopolysaccharide and muramyl dipeptide on food intake. Physi- ology and Behavior, 47: 805-813.

Braz J Med Biol Res 30(2) 1997