The function of anxiety fear related genes in animal models 2

The selectiveCCK2 receptor antagonist L-365,260 not only decreased the anxiety-like behavior in animal models, but also blocked the effect of caerulein in elevated plus maze Mannisto et

is not clearly associated with a single eliciting stimulus, and may last for longer periods of time once activated (Charney and Deutch, 1996)

Some authors argue that external stimuli are insufficient to distinguish fear and anxiety According to them fear is related to action, and particularly to escape and avoidance When the action is blocked, fear is turned into anxiety Therefore fear is an avoidance motive If there were no restraints, internal or external, fear would support the action of fight or flight Anxiety can be defined as unresolved fear, or, alternatively, as a state of undirected arousal following the perception of

threat The alarm or primary anxiety may be channelled or resolved into fear, if

escape is selected as the action option after completely controlled processing of stimulus situation (Lang et al., 2000)

In this work, anxiety will be differentiated from “fear” on the basis of two above cited characteristics First, the two are differentiated by the presence (in fear)

or the absence (in anxiety) of a specific stimulus that elicits the affective response Second, fear is conceptualized as the motivator of a set of coping responses to events

that provide more or less direct threats to the survival or well-being of the organism, whereas anxiety occurs when the situation is too uncontrollable to permit active coping behaviors (Ohman and Mineka, 2001) However, it must be emphasized that anxiety and fear are normal reactions to danger When anxiety and fear are more recurrent and persistent than what is reasonable under the circumstances, and when they impend normal life, an anxiety/fear disorder exists (LeDoux, 1995)

1.2 Anxiety disorders

According to the estimation of ADAS (Anxiety Disorders Association of America, 2002), anxiety disorders are the most common mental illness in the U.S with 19.1 million (13.3%) of the adult U.S population (ages 18-54) affected Characteristics that distinguish abnormal from adaptive anxiety include:

(1) Anxiety out of proportion to the level of threat

(2) Persistence or deterioration without intervention (> 3 weeks)

(3) Symptoms that are unacceptable regardless of the level of threat, including recurrent panic attacks, severe physical symptoms and abnormal beliefs such as thoughts of sudden death

(4) Disruption of usual or desirable functioning

The five anxiety disorders are identified as: Panic Disorder, Compulsive Disorder, Post-Traumatic Stress Disorder, Generalized Anxiety Disorder and Phobias (including Social Phobia, also called Social Anxiety Disorder)

Obsessive-The significance of anxiety/fear in health and disease is well recognized today, but its underlying molecular and neurobiological mechanisms are not well

understood Anxiety/fear is a complex emotional state that cannot be reduced to imbalances of a single neurotransmitter; however, preclinical and clinical evidence has suggested the involvement of noradrenaline (NA), benzodiazepine, serotonin (5-hydroxytryptamine, 5-HT), dopamine neuronal systems in anxiety and fear development And a lot of neurotransmitters are responsible for this process as well Among them, a prominent participation of cholecystokinin (CCK) (Griebel, 1999; Woodruff et al., 1991), 5-HT (Blanchard et al., 1988; Graeff et al., 1996) and corticotropin-releasing factor (CRF) (Eckart et al., 1999; Takahashi, 2001) in anxiety and fear is generally acknowledged today

2 Cholecystokinin (CCK) and its receptors

2.1 CCK: a neurotransmitter

2.1.1 Discovery of CCK and its molecular forms

CCK was first identified and characterized in the gastrointestinal tract as a hormone with a major role in regulating the control of gut motility, pancreatic secretion, and gall bladder contractions In the brain, CCK is also one of the most widely distributed peptides, where it acts as a neurotransmitter CCK fulfills the criteria for a neurotransmitter in the CNS: it is synthesized and stored in nerve terminals and cell bodies, it is released by depolarization, specific receptors and antagonists exist, and it influences the firing rate of other central neurons

This peptide, initially characterized as a 33-amino-acid sequence,is present

in a variety of biologically active molecular formsderived from a 115-amino-acid precursor molecule (prepro-CCK) (Deschenes et al., 1984), such as CCK-58, CCK-

39, CCK-33, CCK-22,sulfated CCK-8 [Asp-Tyr(SO3

(Trp-Met-Asp-Phe-NH2) (Fig 1-1) Several molecular CCK forms, ranging from 4 to 58 amino acids, are found in the brain and peripheral organs The most abundant fragment in the brain is the sulfated octapeptide CCK-8S The tetrapeptide CCK-4 is an important and the shortest biologically active form occurring in the brain as well

2.1.2 Distribution of CCK-related peptides in the central nervous system

CCK is very abundant in the brain, more than in the gut CCK levels are very high (> 4 ng CCK/mg protein) in cerebral cortex, caudate-putamen, hippocampus, and amygdala, while thalamus, hypothalamus and olfactory bulb are lower (1–2 ng/mg protein) The pons, medulla and spinal cord have even lower levels (<1 ng/mg protein), while CCK is barely detectable in the cerebellum (Beinfeld et al., 1981)

The anatomy of central CCKergic projections is very complex and has been the subject of intense investigation since its discovery The ability to visualize CCK mRNA has provided much additional information about the location of CCK cell bodies Many brain regions like the cortex and hippocampus contain both CCK-positive interneurons and a mixture of afferent and/or efferent cells and terminals Certain regions like the striatum, nucleus accumbens and olfactory tubercle have abundant CCK terminals, but few CCK-positive cells

CCK cells are abundant in three sexually dimorphic nuclei in the rat forebrain (the central part of the medial preoptic nucleus, the encapsulated part of the

bed nucleus of the stria terminals, and the posterodorsal part of the media nucleus of the amygdala) CCK mRNA levels are altered by estrogen in these areas

2.2 CCK receptors

2.2.1 Characterization of two CCK receptor subtypes: CCK 1 and CCK 2

Receptors for CCK have been pharmacologically classified on the basis of their affinity for the endogenous peptide agonistsCCK and gastrin, which share the same COOH-terminal pentapeptideamide sequence but differ in sulfation at the sixth

CCKB, have thus been distinguished

A nomenclature committee of the International Union of Pharmacology

receptors, respectively Although many researchers expressed their opposition to this change and felt that the old system was working well (A for alimentary and B for brain), both nomenclatures currently co-exist in the literature This new nomenclature has been adopted in our present study

The CCK1 receptor was first characterized using pancreatic acinar cells (Sankaran et al., 1980), whereas the CCK2 receptor, witha different pharmacological profile, was discovered in the brain(Innis and Snyder, 1980) The gastrin receptor mediatingacid secretion in the stomach was initially thought to constitutea third type

of high-affinity receptor on the basis of its locationand small differences in affinity for CCK and gastrin-like peptides(Song et al., 1993) However, subsequent cloning

of gastrin andCCK-B receptors revealed their molecular identity

CCK1 and CCK2 receptor types have been shown to differ by theirrelative affinity for the natural ligands, their differential distribution, and their molecular structure Both of them have equally high affinity for CCK 8 sulfate, while they differ substantially for unsulfated CCK peptides, gastrin and amidated peptides

subtype, typically found in the pancreas, is relatively specific for sulfated CCK 8; unsulfated CCK 8, CCK 4 and gastrin are 2–3 orders of magnitude less potent For

CCK1 and CCK2/gastrin receptors is tissue dependent

2.2.2 Distribution of CCK receptors

A Distribution in Central Nervous System

Specific CCK-binding sites were demonstrated in membranes from brain homogenates almost two decades ago (Innis and Snyder, 1980; Saito et al., 1980)

hybridization and immunocytochemistry haveinvestigated the regional distribution and specific cellular localizationof CCK receptors throughout the neuraxis Early studies used radioligands such as 125I-CCK-33, 125I-CCK-8, [3H]pentagastrin, [3H]CCK-8, [3H]CCK-4 or [3H]Boc[Nle28,31]CCK27-33 that do not distinguish

CCK-binding sites in several areas, includingthe cerebral cortex, striatum, olfactory bulb

and tubercle, andcertain amygdaloid nuclei Moderate levels were observed in thehippocampus, claustrum, substantia nigra, superior colliculus, periaqueductal gray

hypothalamic nuclei andin the spinal cord

With the advent of specific radioligands that could differentiate between the two types of CCK receptors, it has become apparentthat CCK1 and CCK2 receptors exhibit a sometimes overlapping,yet distinct, distribution throughout the CNS The

receptorsrestricted to rather discrete regions The precise anatomicallocalization of the two CCK receptor types serves to provide morphological substrates for many of the diversefunctions attributed to neural CCK, including involvement in feeding,satiety, cardiovascular regulation, anxiety, pain, analgesia,memory, neuroendocrine

neuropsychiatric disorders(Crawley and Corwin, 1994)

B Distribution in Gastrointestinal and Other Systems

In the gastrointestinal tract and other peripheral systems, CCK1 receptors are present in pancreatic acinar cells, chief cells and D cells of the gastric mucosa, smooth muscle cells of thegallbladder, pyloric sphincter, sphincter of Oddi, some gastrointestinal smooth muscle and enteric neuronal cells, and anterior pituitarycorticotrophs (Wank et al.,1994) CCK1 receptors can also be expressed in several

neuroblastomas,as well as in certain pancreatic carcinoma, neuroblastoma, andlung cancer cell lines Furthermore, CCK1 receptor mRNA has beenfound in esophageal,

gastric, and colon cancers (Clerc et al.,1997) On the other hand, peripheral CCK2receptors are located in smooth muscle cells throughout the gastrointestinal tract (includingthe gallbladder), parietal, enterochromaffin-like, D cells andchief cells of the gastric mucosa, myenteric plexus neurons, pancreaticacinar cells, monocytes, and T lymphocytes Tumors and tumor cell linesexpressing CCK2 receptors include

astrocytomas, andcertain pancreatic and lung cancer cell lines (Reubi etal., 1997)

2.2.3 The molecular biology of the CCK 1 and CCK 2 receptors

In April 1992, the cloning of the rat pancreatic CCK1 receptor (Wank et al., 1992) and the canine parietal gastrin receptor (Kopin et al., 1992) were reported

and CCK2 receptors are 48% identical to each other and code for a protein of about

450 amino acids (Fig 1-2; 1-3) They contain potential sites for N-linked glycosylation and serine phosphorylation Unlike other neurotransmitter receptor systems (e.g., acetylcholine, serotonin and somatostatin), the CCK system, with only two receptor subtypes, appears to be a model of simplicity

mapped to a syntenicregion on chromosome 5 using a wild × inbred backcross panel

fluorescence in situ hybridization The rat CCK1 receptor gene has been localizedto

(Takiguchi et al., 1997)

human/hamster hybrid DNAs (Huppi et al., 1995) Fluorescencein situ hybridization

receptor gene to the distal shortarm of chromosome 11 (11p15.4) The colocalization

of the CCK1 receptor gene withthe dopamine D5 receptor gene at 4p15.1-p15.3 and

pathways by both CCK1 and CCK2 receptors

2.3 Function of CCK-related peptides and CCK receptors

Over the last 20 years, numerous and extensive studies on the functional significance of CCK peptides in the CNS have been published It was demonstrated that biologically active CCK peptides in the brain are involved in the regulation of feeding, the control of learning and memory, behavioral expression of anxiety and panic attacks, mediation of painful stimuli, and modulation of dependence and withdrawal processes as well as functions controlled by the dopaminergic, serotonergic, and opioid systems The referred interactions of CCK with classical

transmission systems have been found in regions both with and without neuronal localization with other transmitters (Crawley and Corwin, 1994)

co-2.3.1 Role in anxiety

The initial suggestion that the CCK system might be involved in anxiety came from experiments of Fekete et al (1981), who reported that injection of CCK8 into the central nucleus of the amygdala enhances behavioral arousal and fear motivation of rats Since this report, evidence from different standard animal models

of anxiety has accumulated to suggest that CCK-related peptides are anxiogenic after peripheral or intracerebral administration Subsequent clinical studies demonstrated that bolus injectionsof the CCK2 receptor agonist CCK4 or pentagastrin provoke panicattacks in patients with panic disorders (Bradwejn et al., 1990) or in healthy

(1994) showed the anxiety-like behavior in Wistar rats in the elevated plus maze and light/dark test after the i.p injection of CCK agonists BC 197 and BDNL Caerulein

is anxiogenic in the elevated plus maze in rats and mice (Mannisto et al., 1994; Singh et al., 1991) CCK8s or CCK8us induced anxiety-like behavior in acoustic startle reflex, elevated plus maze, four-hole box, marble burying test, light/dark test and open field (Biro et al., 1993; Belcheva et al., 1994)

observed by all investigators, and some results have been highly variable and sometimes contradictory The heterogeneity of response produced by CCK administration can be explained by the fact that in some studies, different CCK fragments have been infused in different brain areas in order to delineate the anatomical substrate of CCK-inducing anxiogenic-like effects The local application

of CCK4 in the basolateral amygdala produced an increase in the startle response after acoustic stimulation, while perfusion in the nucleus accumbens did not modify basal startle amplitude (Vaccarino et al., 1997) Studies with CCK8s also yielded a different profile in different regions The local application of CCK8s produced anxiogenic-like effects in the elevated plus-maze when perfusion was performed in the amygdala (Belcheva et al., 1994), posterior nucleus accumbens (Dauge et al., 1990) and dorsal periaqueductual gray matter, but not in the anterior nucleus accumbens (Dauge et al., 1990) Thus, the effects of CCK compounds could vary considerably becauseof existing differences in the CCK fragments, brain areas of

affinity states among species Although negative findings have been obtained in

some anxiety models, it is noteworthy that anxiogenic effects of CCK-like peptide have been reported, in the great part, in different animal species

The hypothesized role for the CCK neuronal system in anxiety is further upheld by studies with CCK antagonists Between the two CCK receptors: CCK1 and CCK2, CCK2 receptors have primarily been implicated in the control of exploratory behavior and the development of anxiety This is proved by the above introduction

administered systemically or intracerebrally, produced anxiogenic-like effects in

(Blommaert et al., 1993; Harro et al., 1993) In addition, these effects could be blocked by CCK2 antagonists, e.g LY288513, which attenuated anxiety behavior in animals when applied alone CI-988 showed anxiolytic effect in conflict test, elevated plus maze, light/dark test and social interaction test (Costall et al., 1991; Bickerdike et al., 1994; Hughes et al., 1990) The selectiveCCK2 receptor antagonist L-365,260 not only decreased the anxiety-like behavior in animal models, but also blocked the effect of caerulein in elevated plus maze (Mannisto et al., 1994) Acute treatment with L-365,260 was also reported to block CCK4-inducedpanic attacks in

demonstrated PD135158 induced the anxiolytic behavior in rats in elevated plus maze In response to the work implicating CCK in anxiety and fear behaviors, recent anxiolytic drug development effects have focused on the therapeutic potential of CCK2 antagonists Potent and selective CCK2 antagonist compounds, such as CI-988

(Park-Davis) and L-365,260 (Merck) have shown considerable promise as novel anxiolytic agents in some standard preclinical (Singh et al., 1991) and clinical tests (Bradwejn et al., 1994)

2.3.2 Role in memory process

There is increasing preclinical evidence that the CCK system may play a role

amygdaloid nuclei, and cerebral cortex It has been suggested that CCK1 and CCK2 receptors have different roles in learning and memory functions In particular, a balance between CCK1 receptor-mediated facilitatory effects and CCK2 receptor-mediated inhibitory effects on memoryretention has been postulated (Lemaire et al., 1992)

agonists in animal models of memory For instance, in the two-trial memory task

with BC 264 has also been described to elicitprominent hypervigilance in monkeys

and Itoh, 1986; Daugé et al., 1992; Lemaire et al., 1992) Factors that potentially contribute to discrepant findings include differences in experimental paradigms,

discrepantfindings on the role of CCK receptors in memory function mightbe due to the heterogeneity of CCK receptors

To date, only a few studies have been devoted to the effects of CCK receptor

electrophysiological investigations of event-related brain potentials showed that ceruletideimproved selective attention in healthy volunteers (Schreiberet al., 1995) Ceruletide has also been reported to improve cognitiveprocessing in young, but not

in elderly, healthy subjects (Dodtet al., 1996) Shlik et al (1998) found that the continuousadministration of the selective CCK2 receptor agonist, CCK-4,producedimpairments in cognitive tests of free recall and recognition, although it had no

may exert a negativeinfluence on memory consolidation andretrieval

3 5-hydroxytryptamine (5-HT) and its receptors

3.1 Introduction

Since its discovery over 50 years ago, 5-HT has been the topic of intense research activity, which has led to the discovery of a range of potential drug targets: the receptors, the metabolising and synthetic enzymes, the uptake sites Some have

been exploited successfully and drugs are available and some have delivered compounds currently under preclinical or clinical evaluation The disorders treated

by drugs modulating 5-HT function cover a wide range, from chemotherapy-induced emesis to depression

The early research on 5-HT was focused on its functions in peripheral tissues, and it was not until much later that its function as a neurotransmitter in the brain was demonstrated This was a major turning point in the discovery of therapeutic agents that modify 5-HT function There are some very important peripheral therapeutic applications of 5-HT research, but it is a fact that a large proportion of the research

to exploit 5-HT pharmacology for therapeutic benefit has focused on its CNS functions

5-HT has been shown to have a multitude of different physiological actions, and this is not surprising given the nature of the 5-HT neuronal system and the variety of different 5-HT receptors (Barnes and Sharp, 1999) The 5-HT neurons originate in the hindbrain in a relatively circumscribed area, but they send projections to most parts of the brain Couple that with the multiple ways that 5-HT can exert its action through many different receptors and it becomes clear why the pharmacology of 5-HT in the CNS is so complex Furthermore, 5-HT is known to interact with other neurotransmitter systems, and the literature is particularly rich in publications in the interactions of 5-HT systems with dopaminergic systems

3.2 5-HT and anxiety

It is widely accepted that 5-HT is involved in the regulation of anxiety; however, there is no agreement on whether 5-HT enhances or, conversely, decreases anxiety The former hypothesis is largely based on earlier experiments in which the effect of drugs acting non-selectively on brain 5-HT mechanisms were measured in animals under punishment or other approach/avoidance conflict situations (Graeff and Schoenfeld, 1970; Tye et al., 1977; Wise et al., 1972) For example, in animal conflict tests (after 24-h water deprivation period, the drinking attempts in rats were punished with an electric shock), drugs or brain lesions that reduce 5-HT output have

a benzodiazepine-like anxiolytic effect Moreover, microinjection of 5-HT receptor antagonists into the amygdala releases punished behavior (increased the number of shocks), whereas 5-HT receptor activation increases response suppression These results support the classical view that 5-HT is anxiogenic In contrast to conflict tests, however, models in which animals actively escape or avoid aversive brain stimulation point to an anxiolytic role for 5-HT Clinical evidence is also controversial For instance, benzodiazepine (BZD) anxiolytics are supposed to alleviate anxiety, at least in part, by decreasing 5-HT release (Wise et al., 1972) However, antidepressant drugs are beneficial in several anxiety disorders when chronically administered Yet, this drug regimen is likely to enhance, rather than impair, 5-HT neurotransmission (Artigas, 1993; Blier et al., 1987)

3.3 5-HT receptor subtypes

To date, 7 different 5-HT receptors and more than 14 subtypes have been identified, some with splice variants and others with isoforms created by mRNA

editing Furthermore, some of the 5-HT receptor isoforms and variants differ in their neuroanatomical distribution There are very few highly selective receptor agonists

or antagonists for these individual receptor subtypes Animal and clinical data have shown that 5-HT1A, 5-HT1B, 5-HT2A and 5-HT3 receptors are all involved in the control of anxiety Identification of many different serotonin receptor subtypes and the development of selective agonists and antagonists for each of these various 5-HT receptor subtypes, have provided the basis for an improved understanding of the dual role of 5-HT in anxiety

3.3.1 5-HT1 receptors

5-HT1 sites have been further subdivided into 1A, 1B, 1C, 1D and 1E subtypes The 5-HT1A and 5-HT1B receptors have been proved to be present in the mammalian brain Numerous clinical and preclinical studies have confirmed the anxiolytic properties of 5-HT1A receptor agonists, such as buspirone (De Vry et al., 1992), which are effective in the treatment of GAD (generalized anxiety disorder)

somatodendritic autoreceptors in midbrain raphe) and postsynaptically (with the highest expression occurring in forebrain limbic structures) (Griebel, 1995; Palacios

et al., 1990) Agonist activation of presynaptic 5-HT1A receptors reduces 5-HT cell firing, synthesis and release (Sinton and Fallon, 1988; Sprouse and Aghajanian, 1987), while activation of postsynaptic receptors typically results in neuronal inhibition (Van den Hooff and Galvan, 1992)

Several agonists, such as 8-OH-DPAT and the slightly less potent pyrimidinylpiperazine derivatives, buspirone, ipsapirone and gepirone, show reasonable selectivity for the 5-HT1A receptor as do a limited number of recently

antagonists display intrinsic activity by themselves, which can potentially complicate interpretation of combined agonist/antagonist effects Nevertheless, `silent doses' (i.e., doses lacking behavioral effects) of 5-HT1A antagonists have been shown to be useful for characterizing the receptor-specificity of 5-HT1A agonist effects (Hogg et al., 1994; Remy et al., 1996)

(Parks et al., 1998; Heisler et al., 1998; Ramboz et al., 1998) Each of the three research groups found a similar spontaneous behavioral phenotype, characterized by elevated anxiety In addition, although all research groups used open-field exploratory behavior as a model for assessing anxiety, two groups, using other behavioral models, the elevated zero or plus maze and a novel object test, confirmed that these mice had elevated anxiety The mechanism of both the increased anxiety and reduced immobility in the antidepressant or stress (forced swim or tail suspension) test models is likely to be a consequence of increased 5-HT availability resulting from the lack or reduction in presynaptic cell body 5-HT1A autoreceptor negative feedback function

A high density of 5-HT1B sites is found in the basal ganglia (particularly the substantia nigra, globus pallidus, caudate putamen, ventral pallidum and entopeduncular nucleus), but also many other regions such as hippocampus and

cortex (Bruinvels et al., 1993) Data from many studies suggest that 5-HT1Breceptors are located on terminals presynaptically and postsynaptically relative to the 5-HT neurons where they play the role of both 5-HT autoreceptors and 5-HT heteroreceptors The former control the release of 5-HT while the latter control the release of non-serotonin neurotransmitters Some reports have suggested the

modulatory role in 5-HT release in this brain region (Moret and Briley, 1997) So far, little information is available concerning the role of 5-HT1B receptor in anxiety-related behaviors Some data has suggested that 5-HT1B has opposite function to 5-

HT1A in animal models

Various agonists that exhibit a certain selectivity for 5-HT1B receptors, such

as RU 24969, TFMPP and eltoprazine, show anxiogenic-like activity in animal models, such as the shock probe conflict procedure, the social interaction test and the elevated plus-maze test both in rats (Pellow et al., 1987) and in mice (Benjamin et al., 1990) By using more selective antagonists, Chopin et al (1998) have shown the possible implication of 5-HT1B receptors in anxiety The 5-HT1B/D and 5-HT1Breceptor antagonists, respectively, GR 127935 and SB 224289, increase the amount

of time spent in the light chamber in the two-compartment paradigm in mice, suggesting the drugs to exert anxiolytic-like activity in this anxiety model

5-HT1B "knockout" mice have been produced by specific ablation of the

5-HT1B receptor gene by homologous recombination Ramboz et al (1996) have found that there is no difference in the level of anxiety between mutant and wild-type mice

as judged by their activity in the light/dark model In contrast, 5-HT1B knockout mice

are less "anxious" compared to wild-type mice in the open-field, the ultrasound vocalisations and the elevated plus-maze (Zhuang et al., 1999) These data in general lend support to the notion that a 5-HT1B receptor antagonist is likely to exert an anxiolytic activity

3.3.2 5-HT2 receptors

To date, three 5-HT2 receptor subtypes have been characterized, namely the 5-HT2A, 5-HT2B, and 5-HT2C receptors The 5-HT2A receptor is predominantly located in the cortex, basal ganglia, and in some components of the limbic system (including hippocampus, septum and amygdala) (Zifa and Fillion, 1992) The highest expression of 5-HT2C receptors is found in choroid plexus, with lower but substantial densities found in parts of the limbic system (as above) and basal ganglia (Wright et al., 1995) Finally, a report suggests that, unlike the 5-HT2A and 5-HT2C subtype, the 5-HT2B receptor has a highly restricted distribution in rat brain, with the highest expression occurring in lateral septum, dorsomedial hypothalamus and medial amygdaloid nucleus (Duxon et al., 1997) The number of 5-HT2 receptor binding sites is decreased in rat frontal cortex following chronic administration of most antidepressant drugs and increased after repeated eletroconvulsive shocks (Sugrue, 1983)

Pharmacologically, few ligands selectively discriminate between the three HT2 receptor subtypes, making it extremely difficult to evaluate their individual functional roles Nevertheless, the combined use of agonists and/or antagonists with varying affinities for each of the 5-HT2 receptor subtypes might permit cautious

5-insight into the functional roles played by each Initial clinical trials using 5-HT2 receptor antagonists (e.g., ritanserin) indicate that they may be useful for the treatment of generalized anxiety disorder, and perhaps some phobias (Taylor et al., 1985) Of the various compounds with antagonist properties at the 5-HT2A receptor tested in the clinic, serazepine (CGS-15040A) showed efficacy in a multicentre trial

in GAD Effects appeared primarily related to the psychic components of anxiety (Katz et al., 1993)

The data regarding the 5-HT2B receptor in anxiety is very rare One report suggested that the injection of 5-HT2B receptor agonist BW 723C86 into the medial amygdaloid nuclei increased the total interaction time of a pair of male rats in the social interaction test The increase in social interaction was prevented by pretreatment with the 5-HT2C/2B receptor antagonist SB 200646A, which did not alter behavior when given alone The results are consistent with the proposal that activation of 5-HT2B receptors in the medial amygdala induces anxiolysis in the social interaction model (Duxon et al., 1997)

More recently, a series of compounds that are selective antagonists at the

be active in animal models of anxiety following acute and chronic administration

receptor antagonist, is currently in late phase II trials for the treatment of anxiety It appeared to have an anxiolytic effect in a variety of tests on animal models, including punished drinking tests, social interaction tests and maze tests in rats and marble burying tests in mice It does not have a muscle relaxant effect but appears to

improve the escape performance of rats in a learned helplessness model, indicating that it may also have potential for the treatment of depression (Pälvimäki et al., 1998)

To investigate 5-HT2C function, 5-HT2C-knockout mice were generated by introducing a nonsense mutation into exon 5 of the X-chromosome gene encoding

performance in the Morris water maze (Tecott et al., 1998) This deficit was not associated with a generalized learning impairment, but rather with a highly selective deficit in long-term potentiation confined to the dentate gyrus and not other hippocampal regions As information processing within the dentate gyrus has also been associated with exploratory behavior, these mice were tested in several models

of exploratory, fear-related behaviors 5-HT2C-/-mice were found not to differ from their normal littermates in contextual fear conditioning, but did exhibit a threefold reduction in emergence neophobia (entry into a brightly lit open field) The data confirms and extends some of the previous findings of the role of 5-HT2 receptors in anxiety

3.3.3 5-HT3 receptors

The highest density of 5-HT3 binding sites is found in the area postrema, with lower but significant densities also found in limbic structures such as the septum, hippocampus and amygdala (Zifa and Fillion, 1992) Agonist activation of the 5-HT3 receptor results in rapid neuronal depolarization, which can be blocked by

a number of selective 5-HT3 antagonists (e.g., tropisetron, ondansetron, GR38032F,

MDL72222) Preliminary clinical trials using 5-HT3 antagonists (e.g., ondansetron and tropisetron) indicate that they may be useful in the treatment of generalized anxiety disorder (Kunovac and Stahl, 1995)

Infusions of ondansetron, granisetron and zacopride into basolateral amygdala produced selective anxiolytic effects in the elevated plus-maze (Tomkins

et al., 1990) as did infusions of GR38032F and tropisetron into central amygdala in the mouse light–dark exploration test (Costall et al., 1989) These latter effects were similar in magnitude to those induced by diazepam The injection of ondansetron, tropisetron, MDL72222, granisetron, and GR65630 also significantly increased rats' social interactions in the high light/unfamiliar test condition, with a magnitude equal

to or greater than that of flurazepam (Higgins et al., 1991) Anxiolytic doses of these 5-HT3 antagonists did not affect locomotor activity in the high light/unfamiliar condition, or social interaction in the less aversive, low light/familiar condition Conversely, selective anxiogenic effects were demonstrated in the light–dark test and social interaction following intra-amygdala infusions of 2-methyl-5-HT (Costall et al., 1989; Higgins et al., 1991) Importantly, the rank order potencies of the 5-HT3 antagonists in the amygdala (i.e., granisetron>tropisetron=GR 65630>ondansetron> MDL72222) agreed well with their relative affinities for the 5-HT3 binding site, and with their relative potencies for inhibiting 5-HT-induced depolarizations of the rat vagus nerve These correlations provide strong evidence that the anxiolytic effects of 5-HT3 antagonists in the amygdala were mediated at the 5-HT3 receptor (Higgins et al., 1991)

On balance, the anxiolytic effects of 5-HT3 antagonists were reasonably consistent across a variety of behavioral tests, although there is some suggestion that 5-HT3 receptors in different regions of the brain might mediate different fear behaviors For example, 5-HT3 antagonists had anti-conflict effects in the hippocampus but not in the amygdala, and increased social interaction in the amygdala but not in the raphe Confirmation of these dissociations may provide important information on understanding the exact role 5-HT3 plays in anxiety and fear development

3.3.4 Other 5-HT receptors

The role of the other 5-HT receptors in anxiety and fear disorders is so far poorly understood At present, there are limitations to pharmacological characterization of 5-HT receptor subtypes in the brain of animals, since agonists, antagonists and antibodies to distinguish all 5-HT receptor subtypes are not yet available

Receptors for 5-HT6, like those for 5-HT4 and 5-HT7, stimulate adenylate cyclase activity This effect can be blocked by antidepressant and antipsychotic drugs such as amitriptyline, clomipramine, clozapine and loxapine in HEK293 cells stably transfected with 5-HT6 receptors (Sleight et al., 1996) The prominent localization of 5-HT6 mRNA in the striatum, nucleus accumbens, olfactory tubercle and substantia nigra, together with its high affinity for both typical and atypical neuroleptics, has led to speculation that this receptor might be one of the target sites

of action for antipsychotic agents In a microdialysis experiment, 5-HT release in the

prefrontal cortex elicited by conditional fear stress (but not the comparable 5-HT release elicited by the control condition, unconditional footshock) was eliminated by the 5-HT6 antisense oligonucleotide osmotic minipump pretreatment regimen (Yoshioka et al., 1998) This data suggests the possibility that 5-HT6 receptors might

be involved in behavioral disorders and in the mechanism of 5-HT-modulating drugs, including antidepressants, which are effective in their treatment

HT7 receptors have been identified in rodent and human brain and, like HT6 receptors, they share high binding affinity for antidepressant and antipsychotic drugs Rat frontal cortical astrocyte 5-HT7 receptors responded to the administration

5-of the antidepressant amitriptyline with an enhanced camp response to 5-HT (Yoshioka et al., 1998) This enhancement was reduced by either a 5-HT7 receptor antisense oligonucleotide or a non-5-HT subtype-selective antagonist, methiothepin

In general, the large amount of 5-HT receptors and subtypes as well as their interaction with other neurotransmitter systems could be responsible for the incredibly diverse actions of serotonin

4 Corticotropin-releasing factor (CRF) and its receptors

4.1 Structure and distribution of CRF-like peptides

CRF exists as a 41-amino-acid polypeptide in a large variety of mammalian species that is synthesized in the hypothalamus and mediates the release of adrenocorticotropic hormone (ACTH) from the anterior pituitary (Spiess et al., 1981; Vale et al., 1981) It is generated by cleavage of the C-terminus of pre-proCRF, the 196-amino-acid precursor (Dautzenberg and Hauger, 2002) 29 different peptides

belonging to the CRF peptide family have been identified in different species Human and rat CRF (h/rCRF) are identical to one another but differ from ovine CRF (oCRF) by seven amino acids (Fig 1-4) Two non-mammalian CRF-like peptides, the 40-amino-acid amphibian peptide sauvagine (SVG) and the 41-amino-acid fish peptide urotensin I (URO), share ~50% sequence identity with h/rCRF Although SVG and URO were originally considered to be CRF homologs in fish and amphibians, the cloning of CRF from fish and frogs and mammalian urocortin (UCN), a 40-amino-acid CRF peptide (Vaughan et al., 1995), established that vertebrates possess additional members of the CRF peptide family To date, UCN has been isolated from human, rat, mouse and sheep (Vaughan et al., 1995; Zhao et al., 1998) UCN is more highly conserved than CRF across species and closely resembles fish URO (53–63% amino acid identity) but diverges from CRF (43–45% amino acid identity) and SVG (35% amino acid identity) The CRF and UCN precursor genes both contain two exons with the second exon encoding the entire precursor protein (Zhao et al., 1998) No additional bioactive peptides appear to be encoded by their precursor genes

CRF mRNA and protein are abundantly distributed in the CNS with major sites of expression in the parvoventricular nucleus of the hypothalamus, cerebral cortex, cerebellum and the amygdalar–hippocampal complex, an area important for stress adaptation, learning and memory (Bittencourt and Sawchenko, 2000) In the periphery, CRF is expressed in the adrenal gland, testis, placenta, gut, spleen, thymus and skin In the brain, expression of UCN mRNA and protein is restricted to the Edinger–Westphal locus, the hypothalamic area and a small population of

neurons in the forebrain In the periphery, UCN is broadly expressed, particularly in the pituitary, gastrointestinal tract, testis, cardiac myocytes, thymus, spleen and kidney

4.2 CRF receptor subtypes

4.2.1 Characterization of two CRF receptor subtypes: CRF 1 and CRF 2

vertebrates as evolutionary distant as fish and humans (Dautzenberg et al., 2001; Arai et al., 2001) Both CRF1 and CRF2 receptors belong to the class B subfamily of

receptor, a 397-438-amino-acid protein, has three functional splice variants, CRF2A–

C Although their N-terminal sequences and tissue distribution differ, these splice variants show no major pharmacological differences There is one functional variant

amino acid identity) The lowest degree of homology (<60% amino acid identity) exists in their extracellular domains, particularly in the N-termini (40% identity) By contrast, intracellular and transmembrane domains are highly homologous (80–85% amino acid identity) with third intracellular loops being completely identical between all cloned CRF receptors (Fig 1-5) Because both CRF1 and CRF2 receptors

probably couples to this intracellular loop A CRF3 receptor isolated from catfish has

identity) Although the pituitary gland is the main locus of CRF1 receptor expression

in many species, CRF3 (but not CRF1) receptor expression has been detected in the

receptor might not exist

4.2.2 CRF receptor expression in the mammalian CNS and periphery

CRF1 receptor mRNA is widely expressed in mammalian brain and pituitary Highest densities of CRF1 receptor mRNA have been found in the anterior pituitary, cerebral cortex, cerebellum, amygdala, hippocampus and olfactory bulbs (Sanchez et al., 1999) In primates, but not in rodents, CRF1 receptor mRNA is found in the hypothalamus and locus coeruleus In the periphery, a low level of CRF1 receptor mRNA expression occurs in the testis, ovary and adrenal gland Although peripheral

inflammation (Uzuki et al., 2001), the CRF1 receptor appears to be more crucial in regulating brain and pituitary functions whereas the CRF2 receptor might be more important in the periphery

CRF2 receptor mRNA is widely expressed in peripheral tissues, particularly

in cardiac myocytes, gastrointestinal tract, lung, ovary and skeletal muscle (Lovenberg et al., 1995) Interestingly, the CRF2B receptor splice variant is the CRF2receptor that is peripherally expressed in rodents whereas the CRF2A receptor is the major splice variant found in the periphery of humans (Dautzenberg et al., 2001) The CRF2A receptor is the dominant CRF2 receptor splice variant expressed in the mammalian brain Central CRF2A receptor mRNA is expressed in a discrete pattern

with highest densities in the parvoventricular nucleus of the hypothalamus, lateral septum, amygdala, hippocampus and retina (Sanchez, et al., 1999; Lovenberg, et al.,

Interestingly, the rat pituitary and cortex express CRF1 receptor mRNA exclusively whereas in primates and Tupaia these tissues express substantial levels of both CRF1and CRF2 receptor mRNA (Sanchez, et al., 1999) Thus, the CRF2 receptor might possess additional functions in primates

4.2.3 Pharmacology of CRF 1 and CRF 2 receptors

Homologs of CRF and CRF-like peptides (e.g UCN, URO and SVG) bind

(Dautzenberg et al., 2001) Only the amphibian CRF1 receptor discriminates between

affinities considerably higher than that of h/rCRF and oCRF, which are more potent

at CRF1 receptors (Dautzenberg et al., 2001) These findings coupled with some data

selectivity SCP/UCNIII and SRP/UCNII were observed to bind exclusively to the CRF2 receptor, although the Kd of SCP/UCNIII is more than tenfold lower than Kdvalues for UCN and SRP/UCNII (Lewis et al., 2001)

The N-terminally truncated SVG analog, antisauvagine-30, has been found to display >300-fold selectivity for CRF2 over CRF1 receptors (Ruhmann et al., 1998)

This CRF2-receptor-specific antagonist has become a valuable tool for in vivo

characterization of CRF2 receptor function Several non-peptidic small molecules have been designed to antagonize CRF1 receptors with >1000-fold selectivity over CRF2 receptors R121919, a water-soluble pyrrolopyrimidine that binds with high affinity to human CRF1 receptors and is well absorbed in humans, has been proved

in clinical trials to produce antidepressant properties in patients with major depression (Zobel et al., 2000)

4.3 CRF-binding protein

A specific CRF-binding protein (CRF-BP) was originally identified by Linton and her associates (1988) and then later purified (Behan et al., 1989) and cloned first in rat (Potter et al., 1991) and then in mouse (Cortright et al., 1995) The protein is likely widespread throughout the vertebrates but remains to be investigated

in detail In mammals, this is a 37-kDa protein with a cysteine disulfide bond arrangement that recalls that of the immunoglobulins Human and rat CRF-BP cDNAs display a high degree of sequence homology and encode a precursor protein consisting of 332 amino acids with one putative N-linked glycosylation site and eleven conserved cysteine residues (Eckart et al., 1999) However, it does not possess a close sequence similarity to known classes of proteins and therefore may represent a prototype of a novel class of proteins An analysis of the ligand structure suggests that the preferable binding motif recognized by the human protein is A-R/K-X-E/Q at positions 22–25 (Sutton et al., 1995) Consequently, sauvagine and ovine CRF do not bind well to the human variant of this protein The following rank

order of ligand affinities was found for CRF-BP: urotensin I > Ucn > h/rCRF > Svg

> oCRF (Kemp et al., 1998)

Expression of CRF-BP in brain is common in all species investigated so far

In the CNS, CRF-BP appears to be partly membrane associated Behan et al (1995) found that ligands that dissociate CRF from CRF-BP increase brain levels of 'free CRF' in Alzheimer's disease (AD) and show cognition-enhancing properties in models of learning and memory in animals without the characteristic stress effects of CRF receptor agonists However, its distribution in the periphery is limited to primates where CRF-BP is localized to the placenta, blood and liver Its original isolation in human plasma led to speculation that its primary function was, in part, binding the large amounts of placental CRF secreted into the blood stream of pregnant women, to protect from over-activation of the adrenal gland Interactions between mammalian corticotropin releasing factor (CRF)-like peptides, their receptors and the CRF binding protein (CRF-BP) were shown in Fig 1-6 To date, the fate of the complex formed by CRF and CRF-BP is not known

exhibit behavioral and physiological symptoms that are associated with overly active CRF-driven stress systems, including comorbid anxiety and agitation and HPA axis dysregulation (Bakshi and Kalin, 2000) Post-mortem measurement of very high concentrations of CRF in the cerebrospinal fluid of severely depressed suicide victims has provided additional support for the hypothesis that chronic hypersecretion of CRF plays a leading role in the etiology of major depression (Arborelius et al., 1999) Moreover, it was reported that Vietnam combat veterans with post-traumatic stress disorder (PTSD), which is characterized by anxiety, flashbacks, and autonomic arousal, show significantly increased concentrations of CRF in CSF (Bremner et al., 1997) and they also exhibit a blunted ACTH response

to CRF (Smith et al., 1989) Researches have also found elevated CSF CRF concentrations in Tourette’s syndrome in which patients show enhanced vulnerability to stress and anxiety (Chappell et al., 1996), and higher concentrations

of CSF CRF during alcohol withdrawal which is characterized by increased anxiety and sympathetic arousal (Adinoff et al., 1996) CSF CRF levels in patients with generalized anxiety disorder are unchanged in comparison with healthy controls (Fossey et al., 1996) Thus, the clinical data provide some evidence for a role of central CRF neuronal systems in anxiety disorders

A lot of evidence also comes from animal studies Centrally administered CRF produces several signs of increased anxiety and transgenic mice that over-express CRF exhibit increased anxiogenic behavior (Stenzel-Poore et al., 1994) Conversely, central administration of either a CRF antisense oligodeoxynucleotide

or a CRF receptor antagonist produce anxiolytic effects in the rat (Skutella et al.,

1994) Microinjections of CRF into the LC decrease open-field activity and increase defensive withdrawal, i.e time spent in a darkened corner of the open-field and an increase in nonambulatory movement (Butler et al., 1990) These behaviors indicate

an increase in anxiety after CRF administration

4.4.2 Role of CRF 1 receptors in anxiety

A number of specific CRF1 receptor antagonists, have been developed One

of the first CRF1 receptor antagonist to be developed and the most extensively tested

in animals is CP-154,526 When administered to rats, this compound produced dependent effects in blocking CRF-induced increases in acoustic startle as well as potentiated startle (Schulz et al., 1996) CP-154,526 was also reported to produce anxiolytic-like effect in elevated plus maze in rats (Lundkvist et al., 1996)

dose-In addition to the rat, the effects of CP-154,526 on mouse emotional behavior were also examined in several measures taken in a light–dark test, a free-exploration test, and stress-induced mouse defense test battery (Griebel et al., 1998) In the hamster, CP-154,526 failed to reverse the submissive–defensive effects of conditioned defeat on subsequent social interactions, albeit plasma ACTH concentrations were reduced in comparison to vehicle-treated animals (Jasnow et al., 1999)

antalarmin is reported to produce deficits in conditioned freezing and blunt the endocrine and autonomic components of the stress response (Habib et al., 2000)

include CRA1000 and CRA1001 (Chaki et al., 1999) These compounds were found

to have anxiolytic-like effects in mice tested in the light–dark test and in rats tested

in the elevated plus maze (Okuyama et al., 1999) DPC904 and R121919 are another two specific CRF1 antagonists, which produce anxiolytic-like behavior (Gilligan et

al., 2000)

behavior suggestive of reduced anxiety In four different tests of spontaneous anxiety––open-field, light–dark box, defensive-withdrawal, elevated plus maze––which normally inhibits behavioral activity, mutant mice consistently exhibited elevated levels of locomotion consistent with an anxiolytic-like profile (Smith et al., 1998; Timpl et al., 1998) In these studies, analyses were also conducted to determine whether there was a general increase in activity following deletion of CRF1 receptors The investigators concluded that because there were no reliable increases in compartment or closed-arm entries made in the light–dark test and in the elevated plus maze, CRF1 receptors play a role in the expression of anxiety without effects on general activity Independent tests conducted on stress-induced hormone

adrenocorticotropin (ACTH) secretion and corticosterone (CORT) after exposure to either restraint or swim stress (Smith et al., 1998; Timpl et al., 1998) Thus, these data provide strong evidence that CRF1 receptors mediate spontaneous anxiety and stress-induced hormone activation

Another strategy to evaluate the functional role of CRF1 receptors is to downregulate its expression using antisense oligonucleotides The administration of

CRF1 antisense oligonucleotides that produced a 10–12% knockdown of brain CRF1receptors induced anxiolytic-like behavioral effects in the defensive-withdrawal test (Heinrichs et al., 1997) In another study, CRF1 antisense-treated rats were less responsive to the anxiety stimulating effects of exogenous CRF than control rats (Skutella et al., 1998) In the elevated plus maze and the open field test, CRF1antisense-treated animals exhibited increased levels of anxiolytic-like behavior In addition, primary pituitary and AtT-20 cells exposed to 72 h of antisense treatment exhibited a reduction in ACTH release after a CRF challenge Receptor

receptor binding after treatment with the CRF1 antisense oligonucleotide

4.4.3 Role of CRF 2 receptors in anxiety

Attempts to determine the functional properties of CRF2 receptors have been facilitated by the synthesis and characterization of a CRF2 specific peptide antagonist called anti-sauvagine-30 (anti-Svg-30) Initial behavioral studies conducted in mice showed that infusion of 400 ng of anti-SVG-30 into the lateral septum had no

antagonist was effective in reversing CRF-induced anxiogenic-like behavior in mice subsequently tested in the elevated plus maze (Radulovic et al., 1999) In this study, anti-Svg-30 delivered to the lateral septum also attenuated immobilization-induced anxiety behavior in the elevated plus maze These behavioral effects in mice may be specific to test conditions involving delivery of the peptide antagonist into the lateral septum

The data from i.c.v injection of anti-Svg-30 is not consistent The anxiety behavior of mouse is increased in the elevated plus maze when anti-Svg-30 (400 ng)

is delivered via i.c.v procedures (Radulovic, et al., 1999), while dose-response studies (Takahashi et al., 2001) conducted in rats suggest that anti-Svg-30 has anxiolytic-like effects similar to those reported for CRF1 antagonists Administration

of anti-Svg-30 using i.c.v methods produced a dose-dependent reduction (1–10 µg)

in anxiety behavior occurring in the conditioned freezing test, the elevated plus maze, and the defensive-withdrawal test Moreover, a dose of anti-Svg-30 (10 µg) that consistently reduced anxiety behavior was further examined in rats tested in a locomotor activity box and found to have no significant locomotor activating effects

knockout animals, CRF2 deficient mice do not show a consistent change in anxiety behavior, either One study reported that CRF2 deficiency produced no significant effects on anxiety responses in the elevated plus maze or in an open field test (Coste

et al., 2000) In another study, CRF2 knockout mice showed no behavioral alterations

in the light–dark test but appeared to exhibit increased anxiety in the elevated plus maze and open field test (Bale et al., 2000) That is, CRF2 knockout mice spent less time exploring the open arms of the elevated plus maze and less time in the inner portion of an open field These anxiogenic-like effects observed in the elevated plus maze and open field may be influenced by the reported increased in CRF mRNA occurring in the central amygdala nucleus of the CRF2 knockout mouse In the rat central amygdala, stress elevates CRF mRNA which is associated with an increase in anxiety behavior

Antisense oligonucleotides that target the CRF2 receptor have also yielded mixed behavioral results In the defensive-withdrawal test, rats administered CRF2antisense oligonucleotides failed to exhibit alterations in anxiety behavior In addition, CRF2 antisense treatment had no effects on anxiety behavior of rats tested

in the elevated plus maze after exposure to swim stress or social defeat (Liebsch et al., 1999) In contrast to these negative results, a recent study demonstrated that CRF2 antisense oligonucleotide treatment attenuated freezing in rats during the immediate post-shock period as well as in a subsequent conditioned testing situation

oligonucleotides may be due to several factors including differences in anxiety testing and potential adverse side effects associated with particular CRF2 antisense oligonucleotides, as suggested by body weight loss

4.5 Role of the CRF system in memory and learning

It has been well recognized today that modulation of learning and memory is one of the major roles of CRF in rodent and human brain Injections of exogenous CRF or displacement of endogenous CRF from the CRF-BP have been employed to study the role of CRF system in learning and memory It was found that intracerebral

training enhances memory in multiple learning tasks, such as visual discrimination learning, inhibitory avoidance learning, spatial learning and fear conditioning (Liang

effect, memory impairment, was observed when CRF was injected

intracerebroventricularly before the memory test (Veldhuis and De Wied, 1984) Increasing evidence suggests that these effects areindependent of the arousal state,

lower doses than the ones that affect arousal, locomotion, or anxiety (Behan et al., 1995) or activate the HPA axis (Diamant and DeWied, 1993).Interestingly, higher

affecting anxiety and food intake(Behan et al., 1995; Heinrichset al., 1997)

Two CRF receptors are believed to be involved in the function of CRF in memory and learning The blockage of CRF receptors using non-selective peptide antagonists seems to attenuate freezing (Kalin et al., 1988; Klain and Takahashi, 1990) However, the underlying mechanisms of CRFR1- or CRFR2-mediated memory behavior as well as the relationship between them have not been well understood The administration of CRF receptor 1 antagonist CP-154,526 suppressed conditioned fear in both rats and mice (Weninger et al., 1999; Hikichi et al., 2000), which confirmed the function of CRF receptor 1 in the impairment of memory The inhibition of brain CRF2 receptor by antisense oligonucleotide also attenuated fear conditioning, indicated by a significant reduction in freezing duration (Ho et al., 2001) This effect was accompanied by a substantial reduction in CRFR2 binding in the lateral septum, which addresses the function of CRFR2 in this brain region in CRF-mediated memory and learning

Radulovic et al (1999) well demonstrated the role of local injection of CRF and CRF receptor antagonists in fear conditioning It was found that injection of

CRF into the dorsal hippocampus before training enhanced acquisition of conditioned fear through CRFR1, as this effect was prevented by the local injection

of the unselective CRFR antagonist astressin, but not by the CRFR2-specific antagonist antisauvagine-30 (anti-Svg-30) In contrast, injection of CRF into the lateral intermediate septum impaired learning through CRFR2, as demonstrated by the ability of antisauvagine-30 to block this effect When antisauvagine-30 was injected alone into the lateral intermediate septum, learning was enhanced However, such effects in learning were not observed when astressin or antisauvagine-30 was injected into the dorsal hippocampus Injection of CRF after the training into the dorsal hippocampus or the lateral intermediate septum also enhanced or impaired learning, respectively Taken together, therefore, these findings indicate that the different behavioral actions of CRF system in learning and memory are mediated through different brain areas and different receptor subtypes

5 Animal Models

5.1 Introduction

The discovery of benzodiazepines (BZs) in the early sixties and their considerable commercial success in the treatment of anxiety have fueled the development of numerous animal models of anxiety Unfortunately, the predictive validity of these initial models has been mainly based on their ability to detect the pharmacological action of BZs, because BZs were the only anxiolytic agents marketed at that time This became evident in the early eighties, when non-BZ

inactive in some anxiety tests, in particular conflict procedures At that time, unconditioned conflict tests such as the elevated plus-maze were developed Later, a second difficulty appeared, when it became evident that anxiety is not a unitary phenomenon but could be divided in various forms including `state' and `trait' anxiety, `normal' and `pathological' anxiety These various forms have been shown

to be differentially sensitive to pharmacological challenge Therefore, when measuring anxiety in animals, it would be useful to have information on the type of anxiety processes which may be involved in a given test Table 1-1 summarizes some animal models of anxiety These models are now extensively used not only to predict the clinical efficacy of pharmacological treatments, but also to phenotype the behavior of transgenic or knockout mice

5.2 Validity criteria of an animal model

What is an animal model of a human behavior? Animal models are experimental preparations developed in one species for the purpose of studying phenomena occurring in another species Kaplan (1973) noted that a model may be valid if it has the same structure as the human behavior or pathology That is whenever a relation holds between two elements of the animal model, a corresponding relation may hold between the corresponding elements of the human behavior Other authors (Treit, 1985) have proposed additional criteria According to them, an animal model should fit predictive validity (pharmacological correlation), face validity (isomorphism) and construct validity (homology and similarity of underlying neurobiological mechanisms) to be suitable for research