Johnson & Johnson Pharmaceutical Research & Development, a Division of Janssen Pharmaceutica N.V.,
Turnhoutseweg 30, 2340 Beerse, Belgium
Abstract: Preclinical and clinical data indicate that corticotropin-releasing factor (CRF) and CRF-related peptides play an important role in stress-related disorders, including psychiatric disorders, such as anxiety disorder, major depression, eating disorders and drug abuse, gastrointestinal disorders, such as irritable bowel syndrome, and immunological disorders, amongst others. Two major CRF receptor subtypes have been identified (CRF, and CRF2, with its prevailing splice variants CRF2o, and CRF2p), which differ in their pharmacology and expression patterns. The recent discovery of selective small-molecule, non-peptidergic CRF! antagonists and of peptidergic CRF2 agonists and antagonists has broadened our understanding of the role of CRF and related peptides in physiological and pathophysiological processes, and opened novel avenues for the development of innovative pharmacological approaches to treat these stress-related disorders, including anxiety and depression, which will be the focus of this chapter.
The 41-amino acid polypeptide corticotropin-releas-ing factor (CRF, also named corticotropin-releasing hormone, CRH) is a hypothalamic hormone, which is released from the parvocellular neurones of the paraventricular nucleus (PVN) into the hypophyseal portal vessels. Upon arrival at the adenohypophysis, CRF activates the transcription of the pro-opiome-lanocortin gene and triggers the release of adrenocorticotropic hormone (ACTH) into the general circulation. ACTH in turn activates the release of glucocorticoids from the cortex of the adrenal gland. This cascade, starting at the level of the hypothalamic PVN, relaying at the level of the adenohypophysis (equivalent to the anterior lobe of the pituitary gland), and ending at the level of the adrenal glands, is called the hypothalamic-pituitary-adrenal (HPA)
axis (Fulford and Harbuz, 2004). Thus, CRF controls the function of the HPA axis during basal activity and stress.
Besides being the most dominant trigger of HPA axis activation during stress, CRF also serves neurotransmitter function in the brain, where it modulates, for example, anxiety-related behaviour, cognitive function, food intake, reproductive behaviour, motor function and sleep, and coordinates the behavioural and autonomic changes during stress (Steckler and Holsboer, 1999).
Alterations in CRF activity have been described in a range of neuroendocrine, neurological and psychiatric disorders, including major depressive disorder, post-traumatic stress disorder (PTSD), schizophrenia, and dementia. In depression, an increased number of CRF-immunoreactive neurones has been reported at the level of the PVN (Raadsheer et al, 1994) and in situ hybridization revealed markedly elevated CRF mRNA levels in the PVN of depressed patients (Raadsheer et al, 1995). An increased CRF-like immunoreactivity has been documented in the
*Fax: +32 1 460 6121; E-mail: tsteckle(á;prdbe.jnj.com cerebrospinal fluid (CSF) of depressed patients (Nemeroff et al, 1984; Banki et al„ 1987; Wong et al, 2000), which seems to decrease upon clinical treatment response to antidepressant medication (DeBellis et al, 1993; Heuser et al, 1998), while lack of normalization of CSF CRF levels during antidepressant treatment may predict early relapse (Banki et al, 1992). Functionally, depressed patients have been reported to have a blunted HPA axis activation to exogenous CRF challenge (Gold et al, 1986; Von Bardeleben and Holsboer, 1988) and to show an abnormal response to combined dexametha-sone/CRF challenge, which has been reported in up to 80-90% of patients (Heuser et al, 1994). Moreover, a decrease in CRF binding sites has been measured in the frontal cortex of suicide victims (Nemeroff et al, 1988), possibly secondary to elevated CRF levels. Indeed, CRF has been demonstrated to downregulate CRF] receptor binding in cortical areas of rats (Brunson et al, 2002), which would provide some support for this idea. Moreover, a recent study showed that expression of CRF) receptor mRNA, but not of CRF2 mRNA, is downregulated in the frontal cortex of depressed patients (Merali et al, 2004). More recently, a shift in the ratio of CRF receptor subtype mRNA expression was seen at the level of the pituitary gland of suicide victims (Hiroi et al, 2001). Thus, there is substantial evidence suggesting abnormal CRF activity in depression, or at least in a subgroup of depressed patients (in particular of the melancholic type; Kasckow et al, 2001), which renders the CRF system an interesting target for the development of new antidepressant drugs (Holmes et al, 2003).
Moreover, a substantial number of animal data point to an important role of CRF in the mediation of anxiety and the regulation of food intake (Steckler and Holsboer, 1999). Clinically, an abnormal response in the combined dexamethasone/CRF test can be seen in panic disorder (Schreiber et al, 1996). In PTSD, the CRF system is also hyperactive (Bremner et al, 1997; Baker et al, 1999; Kasckow et al, 2001). In this respect it is interesting to note that CRF induces kindling in animals, which serves as a model for PTSD (Weiss et al, 1986). Another psychiatric disorder where elevated cerebrospinal fluid CRF levels can be observed is anorexia nervosa. Here, CRF levels have been reported to return to normal level after recovery of body weight (Kaye et al, 1987). This suggests that drugs acting at the CRF system could be useful not only for depression, but also for the treatment of anxiety disorders, eating disorders, as well as for other stress-related disorders (Holmes et al, 2003).
Two main CRF receptor subtypes mediate the effects of CRF-related peptides
Two CRF receptor subtypes can be found in the brain, CRF, and CRF2 (Hauger et al, 2003). These receptors share approximately 70% sequence homology, are class II seven-transmembrane domain G-protein-coupled receptors (GPCRs) and are both positively coupled to adenylate cyclase. Different behavioural functions have been proposed for CRFj and CRF2, with the CRFi receptor being involved in explicit processes, i.e., with the more cognitive aspects of behaviour, including attention, executive function, the conscious experience of emotion, and possibly learning and memory, while the CRF2 receptor may primarily influence implicit processes necessary for survival, i.e., with motivational types of behaviour, including feeding, reproduction and defence (Steckler and Holsboer, 1999).
The human CRFi gene is located on chromosome 17q 12-22 (Polymeropoulos et al, 1995). Several splice variants have been reported for CRF] (CRFla, CRF,p, CRFlc, CRFld, CRFle, CRFlf, CRFlg, CRF,h; Grammatopoulos and Chrousos, 2002), but the functional role of most of these splice variants is unclear at the moment. CRF]o[ seems to be the main receptor splice variant, which is widely expressed in brain and periphery (Chen et al, 1993), although CRF]C has a'so t>een cloned from human brain (Ross et al, 1994). Whether the CRF[ gene is a candidate gene influencing the liability to develop an affective disorder remains to be shown. However, chronic, but not acute, treatment with the tricyclic antidepressant drug amitriptyline reduced CRF! mRNA expression in rat amygdala (Aubry et al, 1997), which links this CRF receptor subtype to antidepressant activity.
The CRF2 receptor is expressed on human chromosome 7pl5-21 (Meyer et al, 1997). Three splice variants have been reported for this CRF receptor subtype in the human, CRF2a, CRF2p and
CRF2y, which only differ at the N-terminus (Grammatopoulos and Chrousos, 2002). Because of this high-sequence similarity, it can be predicted that it would be very difficult to develop compounds with specificity for only one of the CRF2 receptor splice variants. CRF2a is the main CRF2 receptor splice variant with neuronal expression, while CRF2p is primarily expressed by non-neuronal cells in the choroid plexus and cerebral arterioles (Lovenberg et al, 1995). CRF2y has only been found in the human brain, but not in other mammals (Kostich et al, 1998). In the rat, a short isoform of CRF2a (CRF2o,_tr) has been isolated from brain, but binding of CRF does not lead to cAMP accumulation (Miyata et al, 2001), suggesting this to be a silent receptor. A recent study failed to reveal any polymorphism or mutation in the CRF2 gene in a group of depressed patients (Villafuerte et al, 2002) and both acute and chronic treatment with amitriptyline failed to affect amygdaloid CRF2 mRNA level in rats (Aubry et al, 1997), rendering it less likely that the CRF2 receptor plays a major role in the pathogenesis of depression.
A third CRF receptor (CRF3) has been identified in catfish and is approximately 90% homologous to mouse CRF] (Arai et al, 2001), but has not been demonstrated in mammals yet. Moreover, a CRF binding protein (CRF-BP) has been described, a 37kDa glycoprotein that circulates in blood and is also expressed as a membrane protein in the brain. CRF-BP has been shown to function as an endogenous buffer for CRF and related peptides (Potter et al, 1992; Behan et al, 1995), but might have additional function as a modulator of neuronal activity (Ungless et al, 2003). More recently, a possible involvement of the CRF binding protein gene in the genetic vulnerability for depression has been reported (Claes et al, 2003).
Both CRF, and CRF2 receptors bind CRF, but affinity of human CRF for the CRFi receptor is about ten-fold higher than for the CRF2a or CRF2p receptors (Donaldson et al, 1996). In addition to CRF, a number of other endogenous peptidergic ligands have been identified in mammals, including Urocortin I, Urocortin II and Urocortin III, stresscopin (SCP), which can be regarded the human homologue to mouse Urocortin III (SCP3 18), and stresscopin-related peptide (SRP), which seems to be the human homologue to mouse Urocortin II
(SRP6 38 Vaughan et al, 1995; Hsu and Hsueh, 2001; Lewis et al, 2001; Reyes et al, 2001). These peptides differ in their affinity for CRF, and CRF2 receptors, and for CRF-BP: neither Urocortin II nor Urocortin III bind to CRF-BP, whereas Urocortin I binds as well as CRF (Donaldson et al, 1996; Lewis et al, 2001). Urocortin I shows higher affinity for the CRF2a and CRF2p receptors than CRF, which led to initial suggestions that this could be the endogenous ligand for this binding site, but it also has higher affinity for the CRF] receptor subtype than CRF itself (Donaldson et al, 1996). In contrast, Urocortin II and Urocortin III selectively bind to the CRF2, but not to the CRF; receptor subtype (Lewis et al, 2001; Reyes et al, 2001), suggesting that in fact these are the cognate ligands for this receptor.
The two main CRF receptor subtypes differ in their expression pattern in the brain, and this expression pattern is species-dependent: In the rodent, CRF! receptor expression is more widespread than the expression of CRF2 and is almost exclusively observed at the level of the corticotrophs of the anterior pituitary, in frontal cortical areas, the cholinergic basal forebrain, in brainstem cholinergic nuclei, superior colliculus, the basolateral amygdaloid nucleus, cerebellum, red nucleus and the trigeminal nuclei. Strong immunoreactivity for CRF! was also observed at the level of the noradrenergic locus coeruleus and the dopaminergic substantia nigra and ventral tegmental area. CRF2 is more strongly expressed in the PVN, the ventromedial hypothalamic nucleus, the lateral septum, the cortical and medial nuclei of the amygdala, and the serotonergic raphe nuclei. CRF2 mRNA has also been demonstrated in the dopaminergic ventral tegmental nucleus (VTA; Ungless et al, 2003). Low level of CRF2a expression can also be observed in the gonadotrophs of the anterior pituitary, which is sensitive to modulation by stress and glucocorticoids, suggesting a role of CRF2 in the mediation of stress effects on gonadotrophin function at this level (Kageyama et al, 2003). Mixed receptor populations can be observed for the olfactory bulb, the hippocampus, the entorhinal cortex, the bed nucleus of the stria terminalis (BNST), and the periaqueductal grey (PAG) (Chalmers et al, 1995; Lovenberg et al, 1995; Van Pett et al, 2000; Sauvage and Steckler, 2001). More recently, the expression of presynaptic CRF2 receptors has been suggested at the level of vagal afferent terminals (Lawrence et al, 2002). A possible autoreceptor function for CRF2 has also been suggested in the PVN, where the expression of CRF2 mRNA coincides with the cellular distribution of CRF mRNA (Gutman et al, 2001).
In the rhesus monkey, CRF2 is more broadly expressed and found in the neocortex, especially in limbic regions such as prefrontal and cingulate cortices. The expression pattern in the monkey also differs in the amygdala from that in the rat, with highest levels being found in the monkey central nucleus of the amygdala, followed by the medial and basal amygdaloid nuclei. Furthermore, CRF2 has been found in the monkey anterior pituitary (Sanchez et al, 1999), which contrasts the rodent situation, where CRF2 is mainly expressed in the posterior lobe (Van Pett et al, 2000). As in rodents and monkeys, CRF, is the predominant receptor in the human brain, but as in monkey, both receptor subtypes are expressed in the pituitary (Hiroi et al, 2001). These species differences open the possibility that the role of the CRF2 receptor varies across species. For example, the expression of CRF2 in the pituitary suggests that the HPA axis activation in human and nonhuman primates can be maintained through both CRF | and CRF2 receptor subtypes, while it is the CRF | receptor subtype that mediates this response at corticotroph level in the rat and mouse.
In the periphery, CRF] is expressed in skin, spleen, synovial tissue (under arthritic conditions, but not in normal tissue), in the adrenals, and in the placenta and uterus (Grigoriadis et al, 1993; Florio et al, 2000; Willenberg et al, 2000; Murphy et al, 2001; Pisarchik and Slominski, 2001; Zouboulis et al, 2002). CRF2 is found in the cardiovascular system (heart and blood vessels), lung, skeletal muscle, gastrointestinal tract, epididymis, and also in the placenta (Perrin et al, 1995; Florio et al, 2000). Moreover, both CRF[ and CRF2 receptors are expressed in mouse adrenal cortex (Muller et al, 2001), suggesting that some of the effects of CRF-related peptides or of non-peptidergic compounds on HPA axis activity could be mediated directly at this level. This peripheral distribution pattern and the known effects of CRF on peripheral function open the possibility of therapeutic intervention in, for example, stress-induced immune suppression and inflammation, dermatological indications, cardioprotection to stress, irritable bowel syndrome, and uterine contractility. On the downside, this also implies that CRF] or CRF2 compounds developed to treat CNS disorders will likely affect peripheral functions as well and side effects may be seen in the organ systems mentioned above.
Evaluation of CRFi and CRF2 receptor function using mouse mutants
Before reviewing the effects of selective CRF, and CRF2 antagonistic drugs, the lessons learned from mouse mutants with alterations in CRF system activity, especially with changes in CRF! and CRF2 receptor activity, will be briefly discussed.
Transgenic mice overexpressing CRF show an increase in anxiety-related behaviour (Stenzel-Poore et al, 1994; Van Gaalen et al, 2002b), a decrease in sexual activity (Heinrichs et al, 1997) and impaired attentional mechanisms (Van Gaalen et al, 2003). Moreover, these mice exhibit a reduced responsiveness in plasma corticosterone, but normal hypothermic responses, following 5-HT1A receptor challenge (Van Gaalen et al, 2002a), features also reported in depression (Meitzer and Maes, 1995). Although CRF overexpressing mice show a reduction in immobility in forced swim (Van Gaalen et al, 2002b) - the opposite pattern, i.e., increased immobility, might have been expected in this screening test predictive of pharmacological antidepressant-like activity - it is of note that the antidepressant drug Citalopram attenuated the increased anxiety-related behaviour seen in these mice (Van Gaalen et al, 2004).
More recently, another CRF overexpressing mouse has been developed which expresses CRF under the control of a Thy-1 promoter (which should limit overexpression to the nervous system). These mice show an abnormal response in the dexa-methasone suppression test (Groenink et al, 2002). Taken together, CRF overexpression resembles a number of signs and symptoms seen in depression.
In contrast, CRF, receptor knockout mice display reduced anxiety-related behaviour and impaired stress-responsivity (Smith et al, 1998; Timpl et al, 1998), which is in line with previous reports showing anxiolytic-like effects of CRFi antisense oligodeoxy-nucleotide infusions into the lateral ventricles or directly into the amygdala (Liebsch et al, 1995,1998). Similar changes in anxiety-related behaviour have been reported in conditional knockout mice with inactivation of CRFi in the forebrain, suggesting that the behavioural alterations are independent of the activity of the HPA axis (Muller et al, 2003). CRF2 receptor knockout, on the other hand, resulted in anxiogenic-like effects in one study (Bale et al, 2000), gender-specific anxiogenic-like effects (Kishimoto et al, 2000) or failed to display changes in anxiety-related behaviour (Coste et al, 2000). More recently, an increase in depression-like behaviour in the forced swim test has been reported in CRF2 deficient mice, possibly due to a relative overactivity of the CRFi receptor in these mice (Bale and Vale, 2003).
Initiation of the stress response appeared normal in CRF2 knockout mice, but an early termination of the ACTH response was seen, suggesting a role of CRF2 in maintenance of the stress response (Coste et al, 2000). Thus, knockout data suggest that at least part of the behavioural effects of CRF, in particular the anxiogenic and other stress-related properties, are mediated through activation of CRF! and that blockade of the CRFt receptor could have anxiolytic and possibly also antidepressive effects, i.e., the CRFi receptor would represent an interesting target for the development of novel anxiolytic and antidepressant drugs. Conversely, the limited evidence available from knockout studies would suggest that activation of the CRF2 receptor could have protective effects and that activation of CRF2 would be anxiolytic. Indeed, an antiparallel stress system, mediated by CRF2, has been proposed, which should counterbalance the stress-response induced by CRFi activation (Skelton et al, 2000). As such, the CRF2 receptor has been suggested to 'dampen' the activity of CRF] receptor function, both in terms of its effects on anxiety-related behaviour and HPA axis activation, in addition to possible independent anti-stress and anxiolytic functions of CRF2 (Bale et al, 2002). However, there is increasing evidence for an anxiolytic role of CRF2 blockade, as will be discussed below.
Non-peptidergic CRF, antagonists
The first small molecule, non-peptidergic CRF| antagonists were shown in a patent from Nova
Pharmaceuticals in 1991 (Nova Pharm. Corp.: US5063245, 1991). Since then, a number of non-peptidergic CRFi antagonists with biological activity, such as anxiolytic-like and antidepressant-like activity, have been reported, including CRA1000 and CRA1001 (Okuyama et al, 1999; Takamori et al, 2001 a,b); CP 154,526 (Arborelius et al, 2000; Griebel et al, 1998; Kehne et al, 2000; Lundkvist et al, 1998; Millan et al, 2000; Okuyama et al, 1999; Schulz et al, 1996; Takamori, 2001b), antalarmin (Deak et al, 1999; Habib et al, 2000; Zorrilla et al, 2002a,b), DMP695 (Millan et al, 2001), DMP696 (Maciag et al, 2002), DPC904 (Ho et al, 2001), NBI27914 (Smagin et al, 1998; Bakshi et al, 2002), R121919 (Keck et al, 2001; Heinrichs et al, 2002), R278995/ CRA0450 (Chaki et al, 2003) and SSR125543A (Griebel et al, 2002). Structurally, the compounds published so far are closely related and consist of a central, mostly monocyclic, bicyclic or tricyclic, scaffold (in many cases a pyridine or [pyrollo- or pyrazolo-] pyrimidine structure), coupled to an amine (Fig. 1). These compounds show high affinity and selectivity for the CRF! receptor, with IC50's ranging between 5 and 100nM for CRFi, while they lack affinity for the CRF2 receptor, in general with approximately 1000-fold selectivity. They seem to inhibit the CRFi receptor by binding to a transmembrane domain, which suggests an allosteric mechanism of inhibition (Liaw et al, 1997; Nielsen et al, 2000). Two studies reported lack of binding of non-peptidergic CRF] antagonists to CRF-BP (Ardati et al, 1998; Gully et al, 2002), suggesting that non-peptidergic CRF! antagonists are not inactivated by this mechanism and also do not compete with CRF or Urocortin I for this binding site.
Functionally, CRFi antagonists have been shown to inhibit CRF-induced cAMP formation and CRF-stimulated ACTH production in vitro, thereby confirming their antagonistic properties (Keck et al, 2001; Gully et al, 2002; Heinrichs et al, 2002; Chaki et al, 2003; Million et al, 2003). Solubility and bioavailability varies across compounds, but in general has to be considered poor, especially for the earlier compounds, possibly due to their high lipophylicity (Hsin et al, 2002). However, it has clearly been demonstrated that at least some of them, such as CP 154,526, R121919, R278995/CRA0450,
Was this article helpful?
This guide Don't Panic has tips and additional information on what you should do when you are experiencing an anxiety or panic attack. With so much going on in the world today with taking care of your family, working full time, dealing with office politics and other things, you could experience a serious meltdown. All of these things could at one point cause you to stress out and snap.