We have evaluated the feasibility of using food intake and fecal output as markers for ileus in a conscious mouse model by comparing the effect of various insults resulting in ileus on these endpoints. Low doses of LPS, laparotomy/bowel manipulation, 20% TBSA scald burn, and acute pancreatitis comparably produced a transient anorexic effect and delayed defecation consistent with a clinical picture of ileus. The data indicate that both magnitude and duration of the effects were dose dependent establishing a causal link between the initiating insults and the changes in food intake and fecal output. Each insult was self-limiting and mice appeared to have normal grooming habits, vocalizations, and interactions with cage mates within minutes of treatment or recovery from anesthetic. LPS administration in mice has been shown to increase watery secretions in the bowel and induce diarrhea [17
], either of which would potentially confound our fecal output measurements. However, in our study, we saw no evidence of increased watery secretion and diarrhea was not evident at any LPS dose in which fecal output was measured. A recent study showed that a minimum dose of 10 mg/kg was required to produce a significant increase in watery secretion, as measured by intestinal content mass [17
], whereas the highest dose of LPS used in our study was more than an order of magnitude lower (0.4 mg/kg). It might be argued that the inhibition of food intake was caused by the pain and discomfort associated with the various insults. For endotoxin treatment, the delay in food intake directly correlated with delayed gastric emptying suggesting gastric dysfunction was responsible, at least in part, for decreased food intake. Peripheral administration of low doses of LPS is a well established model for sepsis and is associated with the inhibition of food intake in rodents [19
]. In rats, the anorexic effect of LPS appears to be mediated centrally through the activity of serotonin [20
], melanocortins [22
], cytokines [23
], and prostaglandins [25
]. In mice, the anorexigenic effect of peripheral LPS is attenuated by vagotomy [19
] suggesting a brain-gut interaction. There is also evidence that LPS-induced inhibition of food intake in mice is correlated with the inhibition of the orexigenic and motility promoting peptide ghrelin [28
] which is synthesized mainly in gastric tissue and acts centrally. Previous studies have shown that low doses of LPS inhibited gastric emptying [29
]. Here we show that the time course of the LPS induced anorexic effect directly correlated with delayed gastric emptying (Figure ). The parallel recoveries of gastric emptying and food intake suggest that the anorexic effect of LPS was in large part due to gastric dysfunction. Furthermore, although administration of LPS resulted in a significant reduction in movement during the first scotophase, reduced activity alone was not sufficient to explain the anorexic effect (Figure ) as the decrease in the amount of food taken was significantly higher than the decrease in activity. Delayed gastric emptying has also been demonstrated in rodent models of post-operative ileus [1
], thermal injury [33
] and cerulein induced pancreatitis [34
The common effects suggest, at some level, a common mechanism is shared by the multiple insults. One likely candidate is the inflammatory response since endotoxemia [13
], intestinal manipulation [12
], and ischemia/reperfusion injury [39
] have all been shown to promote cellular inflammation within the small bowel and colon in rats. Recently it has been shown in mice that post-operative ileus was associated with inflammatory cell infiltration within the manipulated small intestine, but not the untouched stomach or colon [1
]. In the same study, gastric emptying, measured by scintigraphic imaging, was delayed for 24 hours after insult, recovering within 48 hours. This delay was prevented by inhibition of leukocyte recruitment indicating that the inflammatory cells were responsible for delayed gastric emptying, apparently via inhibitory neural signals since hexamethonium and guanethidine normalized gastric emptying. Inflammatory cell infiltration within the pancreas during cerulein induced acute pancreatitis has been well established, however, it is unclear if similar infiltration within the small bowel muscularis also occurs. Such an event would be confirmatory for the model mentioned above. Our data suggests that inflammatory cell infiltration would be delayed relative to other insults since the effect of higher doses of cerulein did not reach its maximum until the second scotophase. Establishment of inhibitory neural signals is also likely to be sequelae common to the various insults. The role of corticotropin releasing factor (CRF) in both appetite regulation and gastrocolonic motor function has been well established [40
]. CRF receptors are widely expressed in the brain, notably within brain centers that control appetite, the gastrointestinal tract, and on vagal afferent neurons. The known effects of central and peripheral CRF receptors and ligands makes them intriguing investigative targets for systemic control of the different organ systems involved with ileus.
Ileus is most commonly associated with a transient decrease in gastrointestinal motility. In humans, motility resolves differentially within the gastrointestinal tract with the small bowel recovering most rapidly, followed by the stomach and then colon [9
] although it is possible that complete recovery of the entire GI tract is not required for clinical resolution. Food intake and fecal output are the most common clinical markers for resolution of ileus and are more generalized indicators of GI function than measurement of motility, MMC or contractility. As shown in this study, delayed food intake is likely to be modified by behavioral effects and thus provides a more integrated description of gastrointestinal function. In our mouse model, fecal output directly tracked food intake for all insults and doses examined in this study. The fitted models for the 12-hour dose-response curves yielded identical median effective LPS doses for food intake and fecal output providing mathematical verification of this correlation. Since data was collected every 12 hours, it is possible that decreased fecal output lagged food intake by a few hours. Therefore it is difficult to distinguish if this tracking was a function of the initiating insult on colonic activity or normal loss of colonic function due to the fasting caused by the anorexic effects of the insults. The fact that the two endpoints tracked demonstrates the functional coordination of the two organs during ileus and illustrates the necessity for endpoints that examine system wide function in order to develop a complete understanding of the mechanisms of ileus. In the small intestine, 30 minute transit of dye remained unchanged between mice administered endotoxin and control mice 12, 36 and 60 hours after injection. This result indicated that if small bowel transit was compromised by LPS, the effect resolved within 12 hours.
The use of food intake and fecal output as endpoints for ileus in mice revealed dissimilarities from more traditional measures of ileus. In the conscious mouse model described here, food intake and fecal output directly correlated with the extent of bowel manipulation. This result is consistent with at least one other study which showed that, in rats, small bowel smooth muscle impairment and inflammation was directly proportional to the extent of bowel manipulation [37
]. However, several groups have concluded that inhibition of bowel motility was independent of the degree of manipulation or the duration of surgery [9
]. It is probable that the discrepancy in outcomes is due to the different endpoints measured. A large body of research into the mechanisms of ileus focuses on changes in intestinal smooth muscle contractility. In rats, changes in small bowel smooth muscle contractility after endotoxin administration were seen only above the threshold dose of 5 mg/kg [13
]. In our study, food intake and fecal output were responsive to small changes in LPS dose with the lowest dose of LPS tested (0.005 mg/kg) resulting in a significant reduction in both endpoints indicating that they are highly sensitive measures of gastrointestinal function. These data, along with the lack of change in intestinal transit, suggest that low doses of endotoxin produce limited intestinal ileus.
The ability to conveniently monitor the full time course of ileus in the same animal was lacking in most previous models of ileus (although see [44
]). The use of food intake/fecal output as simple, clinically relevant endpoints in un-anesthetized mice should facilitate examination of potential treatment strategies. Improvements in either magnitude or duration of ileus are desirable and treatment strategies would be considered effective if they improved either or both. The fitted model described in this study provides a method for quantitative assessment of both magnitude and duration. Changes in magnitude due to a treatment are most easily assessed by measuring the 12 hour food intake/fecal output in LPS challenged mice. Improvement in treated mice relative to untreated LPS challenged mice would quantitatively assess effectiveness. The specific challenge is not limited to LPS. The 12 hour food intake/fecal output measurements for other insults can be expressed in terms of LPS dose by substituting the amount into the formula for the fitted curves. Thus, in our study, a 20% TBSA burn was comparable to a 0.04 mg/kg LPS dose. Although a more involved assay, duration can be quantitatively monitored by the c-multiplier which will be altered as duration changes. In this scenario, an intervention is used to treat mice challenged with several doses of LPS. An increased c-multiplier, then, would indicate a decrease in the duration of ileus and, thus, indicate an effective intervention. Finally, the ability to monitor ileus without the necessity of harvesting tissue will facilitate gene expression studies which require the same tissue for RNA isolation.