CRF1 receptor signaling pathways are involved in stress-related alterations of colonic function and viscerosensitivity: implications for irritable bowel syndrome


  • Y Taché,

    Corresponding author
    1. CURE/Digestive Diseases Research Center, and Center for Neurovisceral Sciences and Woman's Health, West Los Angeles VA Medical Center, Division of Digestive Diseases, Department of Medicine, University of California Los Angeles, Los Angeles, CA, U.S.A.
    Search for more papers by this author
  • V Martinez,

    1. CURE/Digestive Diseases Research Center, and Center for Neurovisceral Sciences and Woman's Health, West Los Angeles VA Medical Center, Division of Digestive Diseases, Department of Medicine, University of California Los Angeles, Los Angeles, CA, U.S.A.
    2. Integrative Pharmacology-Gastrointestinal Biology, AstraZeneca R&D Mölndal, SE-431 83 Mölndal, Sweden
    Search for more papers by this author
  • L Wang,

    1. CURE/Digestive Diseases Research Center, and Center for Neurovisceral Sciences and Woman's Health, West Los Angeles VA Medical Center, Division of Digestive Diseases, Department of Medicine, University of California Los Angeles, Los Angeles, CA, U.S.A.
    Search for more papers by this author
  • M Million

    1. CURE/Digestive Diseases Research Center, and Center for Neurovisceral Sciences and Woman's Health, West Los Angeles VA Medical Center, Division of Digestive Diseases, Department of Medicine, University of California Los Angeles, Los Angeles, CA, U.S.A.
    Search for more papers by this author

CURE, Bldg 115, Room 117, VA Greater Los Angeles Health Care System, 1130 Wilshire Blvd, Los Angeles, CA 90073, U.S.A. E-mail: ytaché


  • The characterization of corticotropin releasing factor (CRF) and, more recently, the discovery of additional CRF-related ligands, urocortin 1, urocortin 2 and urocortin 3, the cloning of two distinct CRF receptor subtypes, 1 (CRF1) and 2 (CRF2), and the development of selective CRF receptor antagonists provided new insight to unravel the mechanisms of stress. Activation of brain CRF1 receptor signaling pathways is implicated in stress-related endocrine response and the development of anxiety-like behaviors.

  • Compelling evidence in rodents showed also that both central and peripheral injection of CRF and urocortin 1 mimic acute stress-induced colonic response (stimulation of motility, transit, defecation, mucus and watery secretion, increased ionic permeability and occurrence of diarrhea) in rodents. Central CRF enhances colorectal distention-induced visceral pain in rats. Peripheral CRF reduced pain threshold to colonic distention and increased colonic motility in humans.

  • Nonselective CRF1/CRF2 antagonists and selective CRF1 antagonists inhibit exogenous (central or peripheral) CRF- and acute stress-induced activation of colonic myenteric neurons, stimulation of colonic motor function and visceral hyperalgesia while selective CRF2 antagonists have no effect. None of the CRF antagonists influence basal or postprandial colonic function in nonstressed animals.

  • These findings implicate CRF1 receptors in stress-related stimulation of colonic function and hypersensitivity to colorectal distention. Targeting CRF1-dependent pathways may have potential benefit against stress or anxiety-/depression-related functional bowel disorders.

British Journal of Pharmacology (2004) 141, 1321–1330. doi:10.1038/sj.bjp.0705760


corticotropin releasing factor


hypothalamic pituitary adrenal axis




irritable bowel syndrome


locus coeruleus




paraventricular nucleus of the hypothalamus


prostaglandin E2




water avoidance stress


Mammalian CRF ligands

Corticotropin releasing factor (CRF, also known as corticotropin releasing hormone) was characterized by Vale et al. in 1981 as a novel 41-amino-acid hypothalamic peptide that stimulates the synthesis and release of ACTH and β-endorphin from the pituitary (Vale et al., 1981). Since then, activation of CRF signaling pathways in the brain has been shown to reproduce the overall endocrine, autonomic, visceral and behavioral responses to stress (Habib et al., 2000; Jeong et al., 2000; Taché et al., 2001; Bale & Vale, 2004). However, the demonstration that stress-induced behavioral changes occurred in CRF-deficient mice while being abolished by central injection of CRF receptor antagonists suggested the existence of other CRF-related molecules (Weninger et al., 1999). Recently, the CRF family has been expanded by the addition of three novel mammalian CRF-related peptides, urocortin 1 (also known as urocortin), urocortin 2 (also known as stresscopin-related peptide) and urocortin 3 (also known as stresscopin) (Hauger et al., 2003). CRF and CRF-related ligands differ in their brain tissue distribution (Bittencourt et al., 1999; Morin et al., 1999; Reyes et al., 2001; Li et al., 2002) and have a distinct receptor pharmacology (Reyes et al., 2001; Li et al., 2002).

CRF receptors

CRF ligands interact with CRF receptors, subtype 1 (CRF1) and/or subtype 2 (CRF2) cloned from two distinct genes (Perrin & Vale, 1999; Bale & Vale, 2004). Both CRF receptors belong to the class B subtype of G protein-coupled receptors. CRF1 and CRF2 receptors are primary coupled to the Gs-adenylate cyclase signaling pathways (Grammatopoulos & Chrousos, 2002; Bale & Vale, 2004). However, growing evidence indicates that a tissue-specific G-protein coupling can also activate alternative signaling cascades (Grammatopoulos & Chrousos, 2002; Blank et al., 2003). Several CRF1 receptor splice variants, producing at least eight isoforms, α, β, c, d, e, f, g and h, have been identified in humans, several of which are nonfunctional (Grammatopoulos & Chrousos, 2002; Pisarchik & Slominski, 2002). CRF2 receptors are expressed in three functional isoforms α, β and γ, the γ form being found only in human amygdala (Kostich et al., 1998). The CRF2 receptor variants differ in their N-terminal sequence as well as their distribution in both tissues and species, but exhibit identical pharmacological profile (Ardati et al., 1999; Dautzenberg & Hauger, 2002).

One important characteristic of CRF receptor subtypes is their distinct affinity for mammalian CRF family ligands. CRF has higher affinity (10–40-fold) for CRF1 than CRF2 receptor while urocortin 1 binds with equal, and higher than CRF, affinity to both receptors (Vaughan et al., 1995; Lewis et al., 2001; Reyes et al., 2001). Both urocortin 2 and 3 exhibit high selectivity for CRF2 receptor (Lewis et al., 2001; Reyes et al., 2001) (Figure 1).

Figure 1.

Mammalian CRF receptor subtypes 1 and 2, differential affinity of natural ligands on the CRF receptors and synthetic nonselective and selective CRF1 and CRF2 receptor antagonists. For chemical structure, binding affinities and references see Zorrila et al. (2003).

CRF receptor antagonists

Key to the assessment of the role of CRF receptors in the stress response was the development of specific CRF receptor antagonists (Figure 1). Earlier studies relied on the use of nonselective peptide CRF1/CRF2 receptor antagonists, mainly α-helical CRF9–41 (Rivier et al., 1984), D-Phe12 CRF12–41 (Hernandez et al., 1993), followed by the potent and long-acting antagonists, astressin (Gulyas et al., 1995; Miranda et al., 1997) and astressin-B (Rivier et al., 1999). Recently, selective peptide CRF2 receptor antagonists namely, antisauvagine-30 and the more potent and longer-acting analog, astressin2-B were developed (Ruhmann et al., 1998; Higelin et al., 2001; Rivier et al., 2002). The therapeutic potential of CRF1 receptor antagonists for the treatment of neuropsychiatric disorders (Grammatopoulos & Chrousos, 2002) has spurred interest in the development of several nonpeptidic small molecule antagonists that cross the blood–brain barrier. The selective CRF1 receptor antagonists, CP-154,526 (Keller et al., 2002), antalarmin (Webster et al., 1996), NBI 30775 (formerly R-121919) (Heinrichs et al., 2002), NBI-27914 (McCarthy et al., 1999) and recently developed water-soluble compounds, NBI-35965 (Heinrichs et al., 2002; Million et al., 2003) and several others (McCarthy et al., 1999; Gilligan et al., 2000; Zorrila et al., 2003) (Figure 1) have been used to determine the importance of the CRF1 signaling pathways in stress-related activation of the hypothalamic-pituitary–adrenal (HPA) axis and the development of anxiety and depression (Kehne & De Lombaert, 2002; Bale & Vale, 2004). Investigations using mutant mice showed that CRF1 receptor knockout mice have a blunted plasma ACTH and corticosterone responses to stress compared to wild-type littermates. They also display a phenotype of anxiolytic-like behavior and stress hyporesponsivity, further establishing that CRF1 activation is a key component in these responses (Bale & Vale, 2004).

Here, we review growing evidence that supports the role of CRF1 receptors in stress-related alterations of colonic function, and visceral hyperalgesia. These preclinical observations may have implications in the understanding and treatment of irritable bowel syndrome (IBS).

Role of brain CRF1 signaling pathways in acute stress-induced stimulation of colonic motor function

Stress and central injection of CRF and urocortin 1 activate colonic motor function

Experimental studies demonstrated that colonic motility, transit and fecal expulsion are stimulated by acute exposure to a variety of psychological, physical and immune stressors such as open field test, conditioned fear, loud sound, wrap restraint, water avoidance, cold-restraint, inescapable foot or tail shocks and central injection of interleukin-1 (Gué et al., 1991; Enck & Holtmann, 1992; Morrow & Garrick, 1997; Yamamoto et al., 1998; Martinez et al., 2004). Water avoidance (WAS) and restraint, that are largely used as models of acute psychological/physical stress, induce a rapid CRF gene transcription in the parvocellular region of the paraventricular nucleus of the hypothalamus (PVN) associated with the activation of the HPA axis (Lenz et al., 1988a; Ma & Lightman, 1998; Kresse et al., 2001). These stressors also activate sympathetic outflow as monitored by the rise in plasma level of epinephrine and norepinephrine and stimulate sacral parasympathetic pathways (Lenz et al., 1988a; Million et al., 2000). The pattern of colonic motor response to these stress models in experimental animals is relevant to human physiology. Various stressors, including dichotomous listening, intermittent hand immersion in cold water, fear, anxiety and stressful interviews, increased colonic motility in healthy subjects (Rao et al., 1998).

The implication of the brain CRF signaling pathways in the stress-related colonic motor activation has been intensively studied using CRF and CRF-related ligands as well as receptor antagonists (Taché et al., 2001). CRF or urocortin 1 injected intracerebroventricularly (icv) mimicked the colonic response to stress as shown by the increased colonic motility, decreased colonic transit time and induction of defecation in conscious freely moving rats, mice and Monglian gerbils (Williams et al., 1987; Lenz et al., 1988a; Gué et al., 1991; Mönnikes et al., 1994; Miyata et al., 1998; Martinez & Taché, 2001; Okano et al., 2001; Taché et al., 2001; Martinez et al., 2004). The colonic response to icv CRF is centrally mediated (Lenz et al., 1988a). Brain nuclei responsive to CRF have been localized in specific hypothalamic (PVN) and pontine areas (locus coeruleus-LC-, subLC/Barrington nucleus complex) which are also sites involved in CRF-induced behaviors that are symptomatic of anxiety and depression (Mönnikes et al., 1992a; Weiss et al., 1994; Gutman & Nemeroff, 2003). While CRF microinjected into the PVN and LC/Barrington nucleus complex increased colonic motility, decreased colonic transit time, and induced watery fecal output, the peptide had no effect when microinjected into the lateral or anterior hypothalamus under similar conditions in male rats (Mönnikes et al., 1992b; 1993a,1993b; 1994). Conversely, Lewis rats that have a blunted mounting of hypothalamic CRF during stress exposure (Calogero et al., 1992) display a reduced colonic response to WAS associated with a lower activation of PVN neurons as shown by Fos expression (Million et al., 2000).

CRF1 receptor is involved in central CRF- and stress-induced colonic motor stimulation

The central action of CRF is receptor mediated as shown by the dose-related blockade of the colonic motor response to icv CRF by the CRF1/CRF2 receptor antagonists, α-helical CRF9–41, D-Phe12 CRF12–41, or astressin injected icv in rats or mice (Williams et al., 1987; Lenz et al., 1988a; Gué et al., 1991; Martinez et al., 1997; Miyata et al., 1998; Martinez & Taché, 2001; Martinez et al., 2004). Existing evidence indicates that the CRF1 receptor is involved in the colonic motor response to central CRF. The icv injection of CRF1 receptor preferential agonists, ovine CRF and rat/human CRF (Perrin & Vale, 1999) as well as the CRF1/CRF2 agonist, urocortin 1 increased colonic propulsive activity while the CRF2 receptor agonists, urocortin 2 and urocortin 3, injected icv at a similar effective dose than CRF (2 pmol) was inactive in mice (Martinez et al., 2004). In addition, the icv injection of selective CRF1 receptor antagonists, NBI-35965 or NBI-27914, prevented icv CRF-induced stimulation of distal colonic transit and defecation while icv astressin2-B had no effect in rodents (Martinez & Taché, 2001; Martinez et al., 2004).

Studies with selective CRF receptor antagonists and CRF1 receptor knockout mice confirmed the importance of brain CRF1 signaling pathways in the colonic motor response to various stressors. α-Helical CRF9–41, D-Phe12 CRF12–41, and astressin injected icv, into the PVN or LC/Barrington nucleus complex inhibited completely the stimulation of colonic transit and reduced by 50–100% the defecatory response to wrap or partial restraint, WAS, conditioned fear and morphine withdrawal (Table 1 ). Central injection of interleukin-1 β, known to activate hypothalamic CRF release (Turnbull & Rivier, 1999), stimulates colonic motility and the response is blocked by the icv injection of α-helical CRF9–41, (Fargeas et al., 1993). Moreover, selective CRF1 receptor antagonists such as CP-154,526, CRA-1000, NBI-35965, NBI-27914 and antalarmin, injected icv or peripherally, blunted the acceleration of colonic transit induced by restraint, defecation induced by WAS, restraint or social stress (exposure to an intruder) and diarrhea induced by morphine withdrawal (Table 1). In mice, the icv injection of CRF1 receptor antagonist, NBI-35965, abolished the defecatory response to restraint stress whereas the CRF2 receptor antagonist, astressin2-B injected icv at a dose that blocked icv CRF-induced delayed gastric emptying had no effect (Martinez et al., 2004) (Figure 2). Female CRF1 receptor knockout mice showed reduced fecal boli during the open-field test (Bale et al., 2002). Collectively, these data implicate brain CRF1 signaling pathways in the stimulation of colonic motor response to various acute stressors in rodents and monkeys. CRF1 receptors in the PVN are upregulated by central CRF and acute stress (Rivest et al., 1995; Imaki et al., 1996; 2001) and CRF1 receptor antagonists blocked LC activation induced by icv CRF or stress (Schulz et al., 1996). These observations provide additional support for a CRF1 receptor-mediated actions at these sites to stimulate colonic motility.

Table 1. Acute stress-induced activation of colonic motor function: inhibition by pretreatment with CRF antagonists in experimental animals
CRF antagonistsDose/RouteAcute stressColonic inhibitionReferences
  1. ic=intracisternal; icv=intracerebroventricular; IL-1β=interleukin-1β; po=per oral; PVN=paraventricular nucleus of the hypothalamus; sc=subcutaneous; WAS=water avoidance stress. All the studies were done in rats except otherwise stated.

Central CRF1/CRF2receptor antagonists
α-helical CRF9–4150 μg rat−1, icvWrap restraintRapid transit (∼100%)Williams et al. (1987)
α-helical CRF9–4150 μg rat−1, icvPartial restraintRapid transit (∼90%)Lenz et al. (1988b)
α-helical CRF9–4150 μg rat−1, pvnWrap restraintRapid transit (∼100%)Mönnikes et al., 1992b
α-helical CRF9–4150 μg rat−1, pvnWASRapid transit (∼100%)Mönnikes et al. (1993a)
α-helical CRF9–4110 μg rat−1, icvIL-1β, icvSpike burst (100%)Fargeas et al. (1993)
α-helical CRF9–412 μg rat−1, icvConditioned fearSpike burst (100%)Gué et al. (1991)
α-helical CRF9–4150 μg rat−1, icvWASDefecation (60%)Bonaz & Taché (1994)
α-helical CRF9–4150 μg rat−1, icvWrap restraintDefecation (60%)Williams et al. (1987)
α-helical CRF9–4110 μg rat−1, icvMorphine withdrawalDiarrhea (∼50%)Lu et al. (2000)
D-Phe12-CRF12–4110 μg rat−1, icWASDefecation (∼50%)Million et al. (2000)
Astressin10 μg rat−1, icvWASDefecation (56%)Martinez et al. (1997)
    Martinez & Taché (2001)
Selective CRF1receptor antagonists
CP-154,52620 mg kg−1, scPartial restraintTransit (55%)Million et al. (2002)
CP-154,52620 mg kg−1, scWASDefecation (55%)Maillot et al. (2000)
CP-154,52630 mg kg−1, scMorphine withdrawalDiarrhea (∼50%)Lu et al. (2000)
CRA 100020 mg kg−1, scMorphine withdrawalDiarrhea (∼50%)Funada et al. (2001)
NBI-27914100 μg rat−1, icvWASDefecation (67%)Martinez & Taché (2001)
NBI-3596550 μg mouse−1, icvRestraintDefecation (100%)Martinez et al. (2004)
NBI-3596520 mg kg−1, sc, monkeyWASDefecation 53%Million et al. (2003)
Antalarmin20 mg kg−1, poSocial intruderDefecation (40%)Habib et al. (2000)
CRF1 knockout miceOpen field testDefecation (P=0.0008)Bale et al. (2002)
Figure 2.

CRF1 receptor antagonist blocked restraint stress-induced defecation in mice. Groups of mice fed ad libitum were injected icv, under short enflurane anesthesia, with either vehicle, the selective CRF1 antagonist NBI-35965 or the selective CRF2 antagonist astressin2-B. Thereafter, mice were subjected to a 1-h session of stress (restraint in a cylinder) or left undisturbed in their home cages (non-stress). Pellet output was monitored at 15-min intervals for the following 60-min. Mean±s.e.m of cumulative pellet output for the 1-h experimental time. *P<0.05 vs nonstress; #P<0.05 vs vehicle+stress group (ANOVA). Adapted from Martinez et al. (2004, in press).

Central CRF and restraint stimulate colonic motor function through similar autonomic pathways

Convergent studies indicate that central CRF-induced sustained stimulation of colonic motor activity is neurally mediated through the activation of parasympathetic cholinergic pathways and occurs independently from the activation of the HPA axis in rats. Transneuronal labeling showed that the PVN and LC/subLC/Barrington complex send direct projections to the intermediolateral column at the S1 segment of the rat sacral spinal cord that provides parasympathetic innervation of the colon (Vizzard et al., 2000; Taché, 2002). Pharmacological and surgical interventions showed that the colonic motor stimulation (motility, transit, defecation) in response to centrally injected CRF was unchanged after hypophysectomy, adrenalectomy, noradrenergic blockade and by opiate antagonist but attenuated by vagotomy and abolished by ganglionic blockade and atropine (Lenz et al., 1988b; Gué et al., 1991; Mönnikes et al., 1992b; 1993a,1993b). The icv CRF-induced defecation was also blocked by the neurokinin-1 (NK-1) antagonist, TAK-637, and by the 5-HT3 antagonists, ramasetron and azasetron (Miyata et al., 1998; Okano et al., 2001). These data suggest that the parasympathetic activation by icv CRF results in colonic release of substance P, acting on NK-1 receptors, as well as serotonin, activating 5-HT3 receptors. Parallel studies established that pathways that mediate restraint stress-induced colonic motor response are similar to those recruited by icv injection of CRF. The stimulation of colonic transit and defecation by restraint were not modified by hypophysectomy, adrenalectomy, naloxone, indomethacin or bretylium but were attenuated by vagotomy and abolished by ganglionic blockade, atropine and NK-1 or 5-HT3 antagonists (Lenz & Druge, 1990; Castagliuolo et al., 1996; Miyata et al., 1998; Okano et al., 2001).

Role of peripheral CRF receptors in stress-related stimulation of colonic motor and epithelial functions

CRF- and urocortin 1 injected peripherally stimulate colonic function

The role of peripheral CRF signaling pathways in the colonic response to stress has been recently reviewed (Taché & Perdue, 2004). CRF and urocortin 1 injected intraperitoneally (i.p.) or intravenously (i.v.) in conscious rats induced propulsive motor events as shown by the occurrence of clustered spike bursts of long duration in the cecum and proximal colon, reduction of large intestine and distal colonic transit time, defecation and, at the highest doses, watery diarrhea in rats and mice (Williams et al., 1987; Lenz et al., 1988a,1988b; Castagliuolo et al., 1996; Maillot et al., 2000; 2003; Gabry et al., 2002; Martinez et al., 2002; 2004; Million et al., 2002; Saunders et al., 2002a). Likewise, in humans, i.v. CRF increased motility in the descending colon and IBS patients had a greater response than healthy volunteers (Fukudo et al., 1998). In addition to stimulating colonic motor function, i.v. or i.p. CRF increases colonic mucin, rat mast cell protease II and prostaglandin 2 (PGE2) release, ion secretion and permeability to ions, bacterial peptide FMLP, and horseradish peroxidase translocation (Castagliuolo et al., 1996; Santos et al., 1999; Gabry et al., 2002; Saunders et al., 2002b).

CRF1 receptors mediate the peripheral actions of systemic injection of CRF

Although icv and peripheral (i.v. or i.p.) injections of CRF result in a similar pattern of colonic motor activation, several lines of investigation indicate that CRF action is initiated at distinct central and peripheral sites, respectively. First, icv injection of α-helical CRF9–41 did not modify the stimulation of colonic transit induced by i.v. CRF indicating that peripherally injected CRF did not activate brain CRF receptors (Lenz et al., 1988a). Second, i.p. CRF-induced accelerated colonic propulsion is not abolished by ganglionic blockade, unlike after icv administration (Lenz et al., 1988b). Lastly, in the isolated rat distal colon preparation in vitro, CRF added to the bath or through the inferior mesenteric artery increased peristaltic establishing a direct action within the colonic tissue (Mancinelli et al., 1998; Maillot et al., 2000).

The peripheral CRF- and urocortin 1-induced colonic motor stimulation involved activation of CRF1 receptors. The selective CRF2 agonists, urocortin 2 and 3 injected i.v. or i.p. did not alter distal colonic transit or defecation in rats and mice (Million et al., 2002; Martinez et al., 2002; 2004). The i.v. or i.p. CRF-induced stimulation of colonic motor function (clustered spike burst activity, distal transit, defecation and diarrhea) and colonic mucin release were blocked equally by peripheral administration of peptide CRF1/CRF2 receptor antagonists, α-helical CRF9–41 or astressin, which do not cross the blood–brain barrier, and by the CRF1 receptor antagonists, CP-154,526, NBI-35965, NBI-27914 or antalarmin, while the CRF2 receptor antagonists (astressin2-B or antisauvagine-30) had no effect in rats and mice (Maillot et al., 2000; Gabry et al., 2002; Martinez et al., 2002; Million et al., 2002; 2003; Saunders et al., 2002a).

The peripheral mechanisms through which CRF and urocortin 1 influence colonic motor activity may involve a direct activation of colonic myenteric neurons. A robust Fos expression, a marker of neuronal activation, is induced in neurons of the colonic myenteric ganglia by i.p. CRF that is blocked by peripheral injection of astressin and selective CRF1 antagonists (Miampamba et al., 2002) (Figure 3). Double-labeling studies indicate that peripheral CRF activates cholinergic myenteric neurons along with other neurons whose phenotype remains to be further characterized (Miampamba et al., 2002). We recently detected by immunohistochemistry the presence of CRF1 receptors in the rat colonic myenteric and submyenteric neurons (Chatzaki et al., 2004) providing anatomical support for a direct action on colonic enteric neurons. In addition, other studies showed that interstitial cells of Cajal and mast cell degranulation are not involved in the CRF receptor-mediated colonic motor response to i.p. CRF as shown by the use of mast cell deficient KitW/KitW-v mice (Castagliuolo et al., 1998).

Figure 3.

(a) Photomicrographs showing ganglionic Fos immunoreactive (IR) cells (arrows) in the longitudinal muscle/myenteric plexus whole-mount preparations of rat proximal colon labeled with the polyclonal anti-Fos (fos Ab-5). Peripheral injection of CRF (10 μg kg−1, i.p.) induced a robust Fos-expression in the colonic myenteric cells. Pretreatment (−15 min) with peripheral Ast (astressin, 33 μg kg−1, i.p.), or CP (CP 154,526, 20 mg kg−1), blocked the CRF-induced myenteric neurons activation. (b) Number of Fos-IR cells. Each column represents means±s.e.m. of Fos-IR cells labeled with anti-Fos (fos Ab-5). *P<0.001 vs all other corresponding groups; #P<0.05 vs veh/CRF groups (ANOVA followed by Dunn test). Adapted from Miampamba et al. (2002).

Peripheral CRF receptors are involved in stress-related alterations of colonic motor and epithelial function

Activation of colonic CRF signaling pathways is implicated in the colonic responses to acute stress in rodents. Restraint-induced stimulation of colonic transit, defecation and increased motility was prevented or blunted by peripheral injection of α-helical CRF9–41 or astressin but not modified by that of selective CRF2 antagonists (Williams et al., 1987; Castagliuolo et al., 1996; Miyata et al., 1998; Million et al., 2002). A number of CRF1 antagonists given peripherally also inhibits the colonic motor response to various stressors (Table 1). In addition, the peripheral injection of α-helical CRF9–41 blocked the colonic mucosal mast cells activation and PGE2 secretion induced by restraint stress (Castagliuolo et al., 1998). Cold restraint and WAS-induced altered colonic epithelial function (increased ionic secretion and macromolecules permeability) are also blocked by peripheral administration of α-helical CRF9–41 (Santos et al., 1999; Saunders et al., 2002b; Taché and Perdue, 2004). The mechanisms whereby stress recruits the peripheral pool of CRF and/or urocortin 1 to activate peripheral CRF1 receptors in the ensuing colonic motor and epithelial responses are still to be elucidated. Urocortin 1 is present in colonic myenteric neurons and submucosal myenteric plexus as well as in enterochromaffin cells and lamina propria macrophages in rats and humans (Harada et al., 1999; Muramatsu et al., 2000).

Collectively, these data provide conclusive experimental evidence that both central and peripheral CRF receptors are involved in stress-related activation of colonic secretory and motor function (Taché et al., 2001; Taché & Perdue, 2004). It is noteworthy that selective CRF1 antagonists administered peripherally only partially (∼50%) inhibit defecation triggered by various stressors (Table 1). This may indicate submaximal bioavailability of the CRF antagonists at target sites in the brain and periphery, submaximal efficacy at the CRF receptors or both, particularly in view of the fact that the water-soluble and high-affinity CRF1 antagonist, NBI-35965, results in full reversal of restraint stress-induced acceleration of distal colonic transit in mice (Martinez et al., 2004) (Figure 2). It is important to note that neither the peptide CRF1/CRF2 receptor antagonists nor the selective CRF receptor subtype antagonists injected centrally or peripherally influenced basal or postprandial colonic motor function in rodents (Williams et al., 1987; Lenz et al., 1988b; Mönnikes et al., 1992a,1992b; Fargeas et al., 1993; Martinez & Taché, 2001; Martinez et al., 2002; 2004; Million et al., 2003).

Role of CRF receptors in stress-related visceral hyperalgesia

Strains of rats with high anxiety levels displayed enhanced visceral hypersensitivity to colorectal distention (Gunter et al., 2000; Coutinho et al., 2002). Acute or chronic stress facilitates visceral sensitivity to colorectal distention in rats (Bradesi et al., 2002; Coutinho et al., 2002). An initial report by Gué et al. (1997) showed that icv CRF mimicked partial restraint-induced visceral hyperalgesia to colorectal distention in rats, and that the effect of stress was blocked by icv injection of α-helical CRF9–41. Recently, Million et al. (2003) reported that peripheral administration of a selective CRF1 receptor antagonist, NBI-35965, prevented WAS-induced hyperalgesia to colorectal distention in adult rats exposed to neonatal maternal separation that associated with higher brain CRF levels in adults (Cratty et al., 1995; Francis et al., 1999). Likewise, antalarmin injected intraperitoneally prevented both the icv CRF-induced hypersensitivity to colorectal distention as well as the visceral hypersensitivity observed in a high anxiety strain of rats (Cochrane et al., 2001). In healthy humans, stress increased colonic sensation during short-term colonic distention (Ford et al., 1995; Mayer et al., 2001). Peripheral injection of ovine CRF, a preferential CRF1 receptor agonist, reproduced the effects of stress by lowering the stool threshold and sensation of discomfort to distention of the colon (Lembo et al., 1996). A recent clinical investigation showed also that α-helical CRF9–41 injected i.v. significantly reduced the higher symptoms (motility and visceral perception) evoked in IBS diarrhea predominant patients subjected to electrical stimulation compared with controls (Sagami et al., 2004).

Reports so far are consistent with CRF1 receptors being also implicated in stress-related visceral hypersensitivity. Newly developed selective peripherally acting CRF1 receptor antagonists will be useful tools to delineate the respective involvement of central and peripheral CRF1 receptors in the colonic motor and hyperalgesic responses to stress.

CRF1 receptor as a potential therapeutic target for the treatment of irritable bowel syndrome

The clinical relevance of overactive brain CRF signaling pathways in functional bowel disorders, particularly as it relates to IBS, is receiving growing interest. IBS is defined as a functional bowel disorder in which changes in bowel habits, increased sensitivity to colonic distention, and abdominal discomfort/pain are present in the absence of biological markers (Talley & Spiller, 2002). IBS subgroups have been based on the predominance of symptoms: diarrhea, constipation, alternating constipation and diarrhea and abdominal pain (Camilleri, 2001; Talley & Spiller, 2002). In addition to gut symptoms, IBS patients have a high prevalence of coexistent psychiatric disorders, predominantly anxiety–depression and mood disorders. (North et al., 1996; Creed, 1999; Lydiard & Falsetti, 1999; Dunlop et al., 2003). Other studies have shown that stressful life events, including history of major traumatic events in childhood, are important risk factors for IBS and influence the onset and severity of symptoms (Bennett et al., 1998; Kamm, 1998; Solmaz et al., 2003). Conversely, stress-reduction behavioral techniques and psychotropic agents, including tricyclic antidepressants, reduce abdominal pain and bowel symptoms, and the effectiveness of these treatments is best observed in patients in whom diarrhea and abdominal pain predominate (Guthrie et al., 1991; Jones et al., 2000; Camilleri, 2001). Notably, studies in primates and rodents showed that early-life stress is associated with long-standing CRF neuronal hyperactivity in adulthood (Arborelius et al., 1999; Gutman & Nemeroff, 2003). On the other hand, the CRF-suppressive effect of antidepressants has been characterized at the cellular and molecular levels in animals and humans (Arborelius et al., 1999). Antidepressants reduce CRF gene expression at similar brain sites that elicit anxiety and colonic motor responses in rats (Brady et al., 1992; Mönnikes et al., 1993a,1993b; 1994) and decrease elevated CRF concentrations in the cerebrospinal fluid of depressed patients (De Bellis et al., 1993; Keck & Holsboer, 2001).

In summary, preclinical observations established that brain CRF1 receptors convey acute stress-induced anxiogenic behavior and colonic motor stimulation, and recent reports also support their implication in stress-related visceral hypersensitivity. Clinical studies indicate that unrestrained overactivity of brain CRF–CRF1 signaling pathways is related to the psychopathology of anxiety/depression (Bakshi & Kalin, 2000; Keck & Holsboer, 2001). These patients may benefit from treatment with a CRF1 receptor antagonist (Holsboer, 1999) as demonstrated in a small open-label Phase II clinical trial (Zobel et al., 2000). Intravenous CRF also decreased the threshold of pain sensation to colorectal distention in healthy subjects and induced greater stimulation of colonic motility in IBS than in normal subjects (Lembo et al., 1996; Fukudo et al., 1998). These findings support the concept that hyperactivity of the CRF system, presumably the CRF1-dependent pathways, contributes to the comorbidity of anxiety–depression with colonic symptoms in diarrhea predominant IBS patients. Strategies primarily targeted against CRF1 receptors may provide insight into their role in IBS symptoms and a new therapeutic venue for IBS.


The authors' works were supported by the National Institute of Arthritis, Metabolism and Digestive Diseases, Grants RO1 DK-33061, RO1 DK-57238, DK57238-01A1S1, Center Grant DK-41301 (Animal Core) and P50 AR-049550. We thank Drs J. Rivier (Salk Institute, La Jolla, CA), D. Grigoriadis (Neurocrine Biosciences Inc., La Jolla, CA) and E. D. Pagani (Center Research Division, Pfizer Inc., Croton, CT) for the generous supply of different CRF agonists and antagonists used in our studies.