68 MECHANISM OF ACTION OF ANXIOLYTICS JOHN F. TALLMAN JAMES CASSELLA JOHN KEHNE Drugs to reduce anxiety have been used by human beings conditions, including anxiety and depression (3). The link for thousands of years. One of the first anxiolytics and one between CRF and depression is particularly strong, as nu- that continues to be used by humans is ethanol. A detailed merous clinical studies have demonstrated that depressed description of ethanol’s action may be found in Chapter patients showelevated cerebrospinal fluid (CSF) levels of 100. A number of other drugs including the barbiturates CRF, elevated plasma cortisol, and a blunted ACTH re- and the carbamates (meprobamate) were used in the first sponse following intravenous CRF. Successful antidepres- half of the 20th century and some continue to be used sant treatment was shown to have a normalizing effect on today. This chapter focuses on current drugs that are used CRF levels. A role of CRF in anxiety disorders has also been for the treatment of anxiety and approaches that are cur- postulated, though the clinical evidence is not as strong as rently under investigation. it is for depression (4). Preclinical studies have demonstrated that CRF adminis- tered exogenously into the central nervous system (CNS) CORTICOTROPIN-RELEASING FACTOR (CRF) can produce behaviors indicative of anxiety and depression, for example, heightened startle responses, anxiogenic behav- Corticotropin-releasing factor (CRF) is a 41 amino acid iors on the elevated plus maze, decreased food consumption, peptide that plays an important role in mediating the body’s and altered sleep patterns. The anxiogenic effects of CRF physiologic and behavioral responses to stress (1). Figure are not blocked by adrenalectomy, suggesting that they are 68.1 illustrates that this role of CRF may be mediated by centrally mediated effects occurring independently of the multiple sites of action. As a secretagogue, CRF stimulates HPA axis (5). Other studies strengthening the link between the release of adrenocorticotropic hormone (ACTH) from CRF and anxiety include recent work by Kalin et al. (6) the pituitary. In addition, CRF plays a neurotransmitter or demonstrating that a ‘‘fearful’’ phenotype in monkeys is neuromodulatory role through neurons and receptors dis- associated with increased pituitary-adrenal activity and in- tributed in diverse brain regions (2). CRF neurons, localized creased brain CRF levels. Other studies have shown that in the hypothalamic periventricular nucleus, are a major exposure to early postnatal separation stress in rat pups re- mediator of stress-induced activation of the hypothalamic- sults in elevated levels of CRF messenger RNA (mRNA) in pituitary-adrenal (HPA) axis, whereas pathways innervating brain regions including the paraventricular nucleus (PVN) limbic and cortical areas are thought to mediate the behav- and the central nucleus of the amygdala (7,8). ioral effects of CRF. There is a large body of both preclinical and clinical Molecular Mechanism of Action literature implicating a key role of CRF in affective disorders such as anxiety and depression. A significant clinical litera- A substantial scientific effort has been directed toward char- ture suggests that dysfunctions of CRF in its role as a hor- acterizing the molecular biology of CRF pathways (9). Per- mone in the HPA axis or as a neurotransmitter in the brain rin and Vale (9) first isolated CRF and identified it as a may contribute to the etiology of a variety of psychiatric secretagogue for ACTH in primary cultures of rat pituitary cells. CRF activity is shared by two nonmammalian pep- tides, sauvagine and urotensin I, which share a 50% homol- John F. Tallman, James Cassella, and John Kehne: Neurogen Corpora- ogy with CRF, and by a new mammalian peptide, urocortin, tion, Branford, Connecticut. which has a 45% sequence homology. 994 Neuropsychopharmacology: The Fifth Generation of Progress desensitization following exposure to CRF has been demon- strated both in vitro (14) and in vivo (15). Furthermore, chronic stress can down-regulate CRF receptors and de- crease CRF-stimulated cAMP production in multiple brain areas (16,17). Down-regulation of pituitary CRF receptors following adrenalectomy presumably results from decreased ACTH mediated inhibitory feedback, which produces ex- cess CRF stimulation. There are a number of pharmacologic agents available for dissecting the functional significance of CRF-1 and CRF-2 FIGURE 68.1. The role of corticotropin-releasing factor. receptors. Much work has been carried out using the peptide antagonists ␣-helical-CRF (9–41) and D-Phe CRF (12–41). However, these compounds have shortcomings in that they do not penetrate the CNS and therefore have to CRF acts through two Gs-protein coupled receptors, the be administered intracerebrally. Furthermore, they do not CRF-1 and CRF-2 receptor subtypes (9,10). CRF-1 recep- discriminate between CRF receptor subtypes and therefore tors showhomology to a number of other neuropeptide do not allowa determination of their relative contributions receptors, including vasointestinal peptide (VIP) and calci- to behavior. More recently, the development of selective, tonin. Three splice variants of the CRF-2 receptor subtype, nonpeptidic antagonists of the CRF-1 receptor such as CP the CRF-2␣, CRF-2␤, and CRF-2␥, and two splice variants 154,526 (18) have provided important pharmacologic tools of the CRF-1 receptor, have been identified (9). Molecular for the analysis of CRF-1 receptor function. Mutation stud- characterization studies have demonstrated that there is ap- ies have demonstrated that peptide and nonpeptide antago- proximately a 70% sequence homology between CRF-1 and nists bind to different domains of the CRF-1 receptor (9). CRF-2 receptor subtypes. Cloning of the human CRF-2a To date, selective CRF-2 antagonists have not been de- gene revealed that it is 94% identical to the rat CRF-2␣ scribed, though recently, nonpeptide dual antagonists of the receptor and 70% identical to the human CRF-1 receptor. CRF-1 and CRF-2 receptors have been described (19). There is currently no evidence of the existence of the CRF- In addition to CRF receptor subtypes, another potential 2␤ receptor in humans. target for pharmacologic manipulation of CRF is the CRF CRF-1 and CRF-2 receptors have different pharmacol- binding protein (CRF-BP), a 322 amino acid peptide. CRF- ogy and different localizations in the brain and periphery. In BP is a specific carrier for CRF and related peptides found situ hybridization and receptor autoradiography techniques in human plasma and brain. CRF-BP is thought to be a been used to map the relative distributions of CRF-1 and modulator of CRF activity. The CRF-BP is found in high CRF-2 receptors in the rat brain (11,12). High expression densities in the rat amygdala, cortex, and bed nucleus of of CRF-1 receptors was seen in the pituitary, and in a num- the stria terminalis. ber of brain regions including the PVN of the hypothala- mus, cerebral cortex, olfactory bulb, cerebellar cortex, and Genetically Altered Mouse Models basolateral and medial amygdala. In contrast, high densities of CRF-2 are found in more circumscribed regions, includ- Studies utilizing transgenic and knockout mouse models ing the lateral septum, ventromedial nucleus of the thala- have provided important information with regard to the mus, and choroid plexus. Moderate densities of CRF-2 re- contribution of CRF and CRF receptor subtypes to pro- ceptors were reported for the medial amygdala and dorsal cesses including energy balance, emotionality, cognition, raphe nucleus. Further characterization has indicated that and drug dependence (20). This chapter focuses on the evi- the CRF-2␣ splice variant accounts for the brain localization dence implicating CRF and CRF receptors in anxiety states. of CRF-2 receptors, whereas the CRF-2␤ accounts for cho- Overexpression of CRF in transgenic mice produced an- roid plexus. Urocortin, rather than CRF, most closely maps xiogenic effects using either the black-white box test (21) to CRF-2 receptors, leading to the suggestion that it may or the elevated plus maze (22). The latter effect was reversed be the endogenous ligand for CRF-2 receptors. Recent work by central administration of the CRF receptor antagonist has shown that the distribution of CRF-2 receptors may ␣-helical CRF, but not by adrenalectomy, supporting the differ significantly in the nonhuman primate brain relative role of central CRF pathways independent of the HPA axis to the rodent brain such that the CRF-2 subtype may play (22). Studies using antisense directed against CRF in rats a more significant role than previously thought (13). have produced evidence of anxiolytic activity (23). Finally, CRF receptors utilize 3′,5′-cyclic adenosine monophos- overexpression of CRF-BP is anxiolytic, whereas binding phate (cAMP) as a second messenger in the pituitary and protein knockout mice (in which free CRF levels are ele- brain and can be regulated by chronic activation. Thus, vated) display an anxiogenic phenotype in the elevated plus Chapter 68: Mechanism of Action of Anxiolytics 995 maze (24). These data generally support the link between Of the compounds listed above, the compound discov- CRF and anxiety. ered by Neurocrine Biosciences and licensed by Janssen More recently, several studies have highlighted the im- Pharmaceuticals, R-121919, has proceeded the furthest in portance of the CRF-1 receptor subtype in anxiety. CRF- clinical evaluation. A recent report (37) describes results 1 knockout mice demonstrated a diminished anxiogenic re- from a phase II, open-label, dose-escalating trial in which sponse on the elevated plus maze and decreased ACTH 20 severely depressed (Hamilton Depression Score Ͼ25) and corticosterone responses to restraint stress (25). Similar patients were administered R-121919 in one of two dose findings were reported by Timpl et al. (26), using the black- ranges: 5 to 40 mg, or 40 to 80 mg. In the low-dose group, white box anxiety paradigm.