Stress and the colon: central-vagal or direct peripheral effect of CRF?

STRESS, DEFINED AS THE PERCEPTION of threatening stimuli and the emotion and mental activity expended contemplating the threat, has been associated medically with the derangement of normal bowel function since ancient times (Nature of Man, Hippocrates, circa 630 BC). More recently, Bockus (2) provided a cogent description of “motor neuroses of the colon” and noted particularly that the stress and strains of civilization are often reflected in a spastic, hypermotile colon. Bockus (2) speculated that the effect of stress on the colon was probably the result of hyperactivation of vagal projections to the colon. Not long after Bockus, Almy et al. (1) reported experimental observations on the effects of acute mental anxiety in medical students that left no doubt that stressful events can lead immediately to drastic increases in colonic motility. While these earlier efforts established a strong connection between stress, anxiety, and debilitating increases in colonic motility, the responsible physiological mechanisms were not well understood. The discovery and physiological characterization of corticotropin releasing factor (CRF) in the late 70s and early 80s strongly suggested that this brain peptide regulates more functions in response to stress than just the release of adrenocorticotropic hormone from the anterior pituitary (3, 26). CRF, which is released in response to physiological or psychological threat (17), evokes widespread autonomic changes, including depression of gastric motility and decreased colonic transit time (reviewed in Ref. 22). The first suggestion of a link between stress, CRF, and the hypermotile colon was made by Williams et al. (28). These investigators observed that the colonic hypermotility induced by restraint stress in rats could be mimicked by either intravenous or intraventricular CRF injections, and the effect was antagonized by intravenous or intraventricular application of the CRF antagonist, -helical CRF. While Williams et al. (28) established the role of CRF in provoking colonic hypermotility, they did not discriminate the site of action for CRF. Subsequent efforts by several investigators focused on, as Bockus predicted (2), a mechanism in which CRF acts on selected vagal efferent pathways to either inhibit gastric (6, 7, 10, 27) or activate colonic (9, 25) motility. CRF gastroinhibition is primarily via a direct activation of nonadrenergic, noncholinergic (NANC) inhibitory vagal projections (10). Additional evidence suggests that CRF can also induce gastric relaxation by inhibiting gastroexcitatory vagal efferent responses to excitatory TRHergic inputs and to thyrotropin-releasing hormone (TRH) itself (6, 7, 27). The mechanism of action of CRF for this latter effect, however, is not clear, because CRF only produces excitatory responses within identified vagal efferent neurons both via a direct effect on their membrane and, indirectly, via an increase of the excitatory inputs impinging on them (10). Perhaps CRF produces its effects to withdraw excitatory input to the stomach by modulating yet-to-be-discovered transduction mechanisms within gastroexcitatory vagal efferents. In the present issue of Am J Physiol Regul Integr Comp Physiol, a study by Tsukamoto et al. (25) suggests that the stress effects that increase colonic motility are produced exclusively as a consequence of a central action of CRF on vagal excitatory inputs to the colon. These authors recorded colonic motility with a strain gauge to show that restraint stress, as well as intraventricular injections of CRF, produce significant increases in colonic motility. The colonic contractions induced by CRF were eliminated by intracerebroventricular, but not peripheral, pretreatment with the nonselective CRF receptor antagonist astressin, by hexamethonium, atropine, or truncal vagotomy. In vitro, muscle strip preparations of the colon, instead, were insensitive to CRF (25). The data of Tsukamoto et al. (25) are clear enough; there is, however, compelling evidence suggesting that the centralCRF-vagal efferent mechanism may also be paralleled by the direct action of CRF on the colon itself. Several observations support this possibility, including the demonstration of the equal effectiveness of CRF to augment colonic motility, whether the peptide is delivered centrally or peripherally (13, 14, 25, 28). CRF has also been reported to increase colonic electrical activity when delivered to the splanchnic arterial circulation (13). Recent work of Liu and colleagues (11, 12) and Chatzaki et al. (4) have clearly demonstrated the presence of CRF type 1 receptors in the myenteric and submucosal plexus of the gut, including the colon. There is clearly a mismatch between the bodies of work suggesting that CRF operates exclusively on vagal efferent fibers to drive colonic motility (25) or that CRF action is directly on the colon (4, 9, 11–14, 16). It should be noted that subtle variations in the experimental approaches between investigators are, at times, sufficient to explain differences in results. For example, studies showing a direct colonic action of CRF demonstrate increased electrical activity (13) and colonic enteric neuron activation (12, 16) but not changes in colonic motility (25). Strain gauge studies of CRF involvement in stress-related increases in colonic motility only detect the central nervous system-vagal mechanism (25). It is possible that CRF can act peripherally to activate the enteric plexus and to produce changes in smooth muscle electrical activity, but these changes fail to translate into significant motility events. Conversely, it is possible that the narrow muscle strip preparations (3 mm) used by Tsukamoto et al. (25) may not possess sufficient intact myenteric or submucosal plexus to generate a response to CRF. Something else to consider is that while vagotomy eliminates any central effects of CRF (25), this procedure also Address for reprint requests and other correspondence: R. A. Travagli, Dept. of Neuroscience, Pennington Biomedical Research Center, Louisiana State Univ. System, 6400 Perkins Rd., Baton Rouge, LA 70808 (e-mail: [email protected]). Am J Physiol Regul Integr Comp Physiol 290: R1535–R1536, 2006; doi:10.1152/ajpregu.00011.2006.