| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Pharmacology, Bonn University Clinic, Bonn, Germany
 |
Abstract |
|---|
 Top Next |
|---|
|
I. Introduction: Background, Definitions, and Scope |
|---|
|
TopPreviousNext |
|---|
; Hoyer et al., 1994, 2002; Baumgarten and Göthert, 1997; Barnes and Sharp, 1999; Sari, 2004; Bockaert et al., 2006). Seven of these 14 receptors are discussed in the context of presynaptic modulation of ACh, NA, DA, or GABA release and of the role of GABAergic interneurons in the release modulation of these neurotransmitters. The main features, their occurrence in the brain, and their functional and therapeutic implications are summarized in Table 1.
|
|
; Baumgarten and Göthert, 1997). Exceptionally, the serotoninergic axons innervating the cerebellum originate to a major part in the nucleus reticularis paragigantocellularis and nucleus pontis oralis of the brainstem (Bishop and Ho, 1985). As a result of 5-HT release from nerve terminals in virtually all regions of the CNS, the serotoninergic system is involved in many important physiological functions such as the control of blood pressure, body temperature, appetite, release of prolactin and other hormones, perception of pain, and emotional behavior. Accordingly, the serotoninergic system also plays a role in the pathogenesis of diseases in which these functions are disturbed, e.g., hypertension, hormonal dysfunction, depression, and anxiety. In this context, it should also be noted that the serotoninergic system is influenced by clinically effective drugs such as antihypertensives, antidepressants, anxiolytics, and antipsychotics, many of which act at one or more of the 14 receptor types mentioned above. 5-HT receptors may also become targets of newly developed drugs with more favorable properties.
To define the receptors under review, a simple model of three successive neurons should be considered (Fig. 1). The first neuron is serotoninergic ("neuron I") and releases 5-HT at axoaxonal synapses directly to the 5-HT receptors of the terminal boutons/varicosities of the axon of a second, nonserotoninergic neuron ("neuron II," e.g., cholinergic, dopaminergic, or noradrenergic). Alternatively, 5-HT may be released from neuron I terminals into the vicinity of neuron II terminals (Beaudet and Descarries, 1978; Törk, 1990) and reaches the 5-HT receptors on the neuron II terminals by diffusion over relatively long distances. The terminals of neuron II form synapses with "neuron III." Looking back from neuron III, the 5-HT receptors on the neuron II terminals are considered presynaptic or prejunctional. The presynaptic 5-HT receptors on the neuron II terminals belong to the category of heteroreceptors, which by definition are stimulated by neurotransmitters other than those released by the neuron on which the receptor resides. Accordingly, presynaptic 5-HT heteroreceptors can be defined as release-modulating 5-HT receptors on nonserotoninergic varicosities (synaptic boutons) of axon terminals (Fig. 1).
|
; Timmermans and Thoolen, 1987; Middlemiss, 1988; Starke et al., 1989; Göthert and Schlicker, 1997) and presynaptic heteroreceptors on serotoninergic axon terminals (Göthert and Schlicker, 1991). The present article completes the review series on the cross-talk between serotoninergic and nonserotoninergic axon terminals by focusing on presynaptic 5-HT heteroreceptors on nonserotoninergic axon terminals.
However, in many cases the exact location of the modulatory 5-HT receptors, i.e., presynaptically or somadentritically on the directly modulated neuron or on interneurons, is still an open question due to the lack of appropriate in vitro experiments (e.g., 5-HT4 receptors facilitating ACh release; section III.F.). In several in vivo and in vitro investigations, modulation of release by 5-HT receptor ligands was interpreted in terms of direct operation of a 5-HT heteroreceptor on the respective neuron although the primary cell effect, depending on the receptor signal transduction, did not conform (was even opposite) to such an interpretation. Examples are 5-HT3 receptors inhibiting ACh release (section III.E.) and 5-HT1A receptors facilitating ACh release (section III.B.). These results should be reinterpretated. In both cases inhibitory GABAergic interneurons rather than the cholinergic nerve terminals may be assumed to be the site of action of 5-HT. The release-modulating 5-HT receptors involved are located, at least to a major part, on the axon terminals of GABAergic interneurons (Fig. 2). A crucial prerequisite for the indirect modulation of the release of a neurotransmitter via an inhibitory GABAergic interneuron has to be taken into account: the nerves from which this neurotransmitter is released must be endowed with GABA receptors. In fact, evidence has been presented that inhibitory GABAA and/or GABAB receptors are operative on cholinergic (Supavilai and Karobath, 1985; Moor et al., 1998; Ikarashi et al., 1998; Vazquez and Baghdoyan, 2003), dopaminergic (Ronken et al., 1993; Steiniger and Kretschmer, 2003), and noradrenergic neurons (Fung and Fillenz, 1983; Suzdak and Gianutsos, 1985; Tanaka et al., 2002; Sakamaki et al., 2004; Ushigome et al., 2004)
|
Presynaptic 5-HT heteroreceptors play mainly a physiological role in the local fine regulation of the release of the transmitter from the axon terminal on which the receptors are located. They inhibit or facilitate transmitter release in response to action potentials invading the varicosities. In contrast, somadendritic receptors modify the function of the whole neuron with all branches of its axon. Local modulation via presynaptic heteroreceptors is more pronounced as the 5-HT concentration in the biophase of the receptors increases. This concentration depends not only on the serotoninergic neuronal activity but also on the activity of the transmitter transporter. In addition to these modulatory effects, presynaptic 5-HT heteroreceptors can directly stimulate neurotransmitter release; this has recently been demonstrated for, e.g., presynaptic 5-HT3 receptors on GABAergic axon terminals (Turner et al., 2004; Dorostkar and Boehm, 2007; Dorostkar and Boehm, 2005).
This review focuses on 5-HT heteroreceptor-mediated modulation of transmitter release from cholinergic, dopaminergic, noradrenergic, and GABAergic neurons; in particular, 5-HT1A, 5-HT1B, 5-HT2A, 5-HT2C, 5-HT3, and 5-HT4 receptors play a role in modulation of these neurotransmitters (Table 1). A previous article reviewed the literature before 1998 on presynaptic and nonsynaptic interactions between the same transmitters as are the focus of this article (Vizi and Kiss, 1998). It is complementary to this article because it focused on hippocampal transmission only, including the interaction between glutamatergic neurons, GABAergic interneurons, raphehippocampal serotoninergic, coeruleohippocampal noradrenergic, and septohippocampal cholinergic projections. A more recent article reviewed the literature on the modulation of neurotransmitter release via multiple auto- and heteroreceptors in the human brain (Raiteri, 2006). Only a small part of that article is devoted to human 5-HT receptors and, thus, it overlaps only to a minor extent with this review.
|
II. Experimental Approaches |
|---|
|
TopPreviousNext |
|---|
Evidence for the operation of 5-HT heteroreceptors mediating modulation or stimulation of neurotransmitter release has been provided by in vitro experiments in which the effects of 5-HT receptor ligands on the overflow of radioactively labeled or endogenous neurotransmitter from brain or spinal cord slices or synaptosomes has been investigated (for references, see Tables 2, 3, 4, 5, 6, 7, 8, 9, 10, 11). Briefly, CNS preparations were preincubated with the labeled transmitter or its precursor. Subsequently stimulation-evoked overflow of radioactivity from the superfused (or incubated in a few cases) preparations was measured. As a rule, electrical impulses, high K+ or, in the case of excitatory 5-HT3 receptors, 5-HT3 receptor agonists were used for stimulation (see Tables 2, 3, 4, 5, 6, 7, 8, 9, 10, 11). Stimulation-evoked overflow of radioactivity or labeled transmitter, in particular in the presence of an inhibitor of the respective neuronal transmitter transporter, reflects the release of endogenous transmitter. Therefore, in such cases, the term transmitter release is used. Determination of depolarization-evoked transmitter release from superfused synaptosomes, i.e., torn-off and resealed varicosities, is the most appropriate method to prove the presynaptic location of the 5-HT heteroreceptor under study in such functional investigations. An alternative approach is the measurement of K+-evoked transmitter release from superfused slices in the presence of tetrodotoxin (TTX), which blocks the Na+ channel-dependent propagation of action potentials along the axon, thus excluding the involvement of somadendritic receptors.
|
|
|
|
|
|
|
|
|
According to the concept of transmitter modulation via presynaptic inhibitory or facilitatory 5-HT heteroreceptors, 5-HT receptor agonists should inhibit or facilitate, respectively, transmitter release in a manner susceptible to blockade by an appropriate 5-HT receptor antagonist. Experiments on slices also provide an answer to the question whether the receptor is tonically activated by 5-HT released from neuron I of Fig. 1; if so, appropriate 5-HT receptor antagonists, given alone, should produce an effect opposite to that of corresponding 5-HT agonists by blocking the effect of endogenous 5-HT released into the biophase of the 5-HT-heteroreceptor. Such information cannot be derived from experiments on superfused synaptosomes because the superfusion flow effectively removes endogenous 5-HT, released from serotoninergic synaptosomes, from the biophase of 5-HT heteroreceptors on neighboring nonserotoninergic synaptosomes.
Knockout mice have been particularly useful for the identification of a presynaptic auto- or heteroreceptor. If modulation of neurotransmitter release only occurs in wild-type but not in the respective receptor knockout mice, the function of a presynaptic receptor can be convincingly shown. This method has been used to prove the function of the presynaptic 5-HT1B receptor in the mouse hippocampus (Rutz et al., 2006).
5-HT heteroreceptor research experienced a renaissance with the development of more sophisticated procedures comprising cell biological and electrophysiological techniques, which were applied to GABAergic terminals (Table 10). For example, such experiments were performed on isolated single hippocampal CA1 pyramidal neurons with numerous adherent functional presynaptic terminals of GABAergic interneurons which, in turn, are endowed with presynaptic 5-HT heteroreceptors. This preparation is called the "synaptic bouton preparation" (Rhee et al., 1999; Akaike and Moorhouse, 2003). The neurons were isolated using an enzyme-free mechanical procedure. In this preparation voltage clamp recordings from the pyramidal cells were performed, and single boutons were focally stimulated. Administration of the 5-HT1A receptor agonist, 8-OH-DPAT, and the 5-HT3 receptor agonist, m-chlorophenylbiguanide, revealed that all boutons (= varicosities) contained inhibitory 5-HT1A receptors and that a subset of boutons was endowed with both 5-HT1A and excitatory 5-HT3 receptors (Katsurabayashi et al., 2003). Another approach to the identification of presynaptic 5-HT3 heteroreceptors on GABAergic axon terminals is based on the ability of rat hippocampal neurons to form autapses in single neuron microcultures; the function of this neuron can be analyzed by conventional whole-cell patch-clamp recordings from the neuronal soma (Dorostkar and Boehm, 2005, 2007). Using this preparation of hippocampal GABAergic neurons, 5-HT caused rapidly activating inward currents, which were reduced by tropisetron [3-tropanylindole-3-carboxylate hydrochloride (5-HT3 antagonist)], leading to the conclusion that GABAergic hippocampal neurons express 5-HT3 receptors, activation of which causes depolarization and a stimulation of GABA release. This interpretation was supported by experiments on conventional mass cultures of hippocampal neurons in which 5-HT produced a 25-fold rise in the frequency of miniature inhibitory postsynaptic currents in a manner sensitive to antagonism by tropisetron (Dorostkar and Boehm, 2005, 2007).
Attempts have been made to demonstrate the presence of certain presynaptic 5-HT heteroreceptors by modern histochemical and cytochemical techniques. In one type of these experiments, in situ hybridization analyses of the respective 5-HT receptor mRNA was combined with determination of the corresponding binding sites by quantitative autoradiography. When the respective 5-HT binding sites can be identified in certain brain regions but no corresponding mRNA is detectable, the 5-HT receptor under consideration may be assumed to occur as presynaptic receptor. This assumption is based on the fact that these receptors are synthesized in the somadendritic area of the neuron and transported to the terminals which, in contrast to the somadendritic area, contain at best low amounts of the respective 5-HT receptor mRNA. Such investigations are available for only few receptors, among them 5-HT1B and 5-HT1D receptors (Bruinvels et al., 1994a,b). However, suggestions concerning the type of neurotransmitter whose release is modulated at the corresponding nerve terminals are largely speculative; not only heteroreceptors on different types of nerve terminals but also autoreceptors may be visualized. In another approach to identifying presynaptic receptors more directly, synaptosomes have been used. Thus, in synaptosomes prepared from corpus striatum, hippocampus, amygdala, and cerebellum of the rat (Nayak et al., 1999), 5-HT3 receptor protein was detected by immunocytochemistry. In synaptosomes from all of these regions the 5-HT3 receptor immunoreactivity was colocalized with the synaptic vesicle protein synaptophysin, clearly indicating a presynaptic localization of these 5-HT3 receptors. However, the synaptosomal preparation did also not allow assignment of the presynaptic 5-HT3 receptor to certain types of neurons. Such studies, in particular when considering the limitations mentioned, are not the subject of this review article dealing with the release of neurotransmitters.
The most frequently applied in vivo technique has been intracerebral microdialysis in anesthetized or freely moving animals (for references, see Tables 2, 3, 4, 5, 6, 8, and 9). The respective neurotransmitter and its metabolites in the dialysates have been determined by high-pressure liquid chromatography. This technique made it possible to directly apply 5-HT receptor agonists and/or antagonists via the dialysis probe to the brain area under study and to simultaneously measure their influence on the overflow of the neurotransmitter under investigation. In many cases, the receptor ligands have been injected systemically; if not supplemented by other experimental results, such data hardly allow a conclusion as to the location within the CNS of the receptor involved (section II.C.). Two further in vivo methods providing similar informations on drug-induced modification of transmitter release in certain brain regions, i.e., the technique of an epidurally implanted cup on the cerebral cortex (Beani et al., 1968; Siniscalchi et al., 1991) and the push-pull technique (Guan and McBride, 1989), have also been used for the identification of presynaptic 5-HT heteroreceptors in the brain. Microdialysis experiments on receptor knockout mice will provide considerable progress also in in vivo receptor identification and determination of the functional role.
C. Advantages and Limitations of the Techniques Applied
In several conventional superfusion experiments on CNS preparations, the location of a 5-HT heteroreceptor on a certain nerve terminal (neuron II) (Fig. 1) has been unanimously proven by determination of depolarization-evoked release of the respective transmitter from slices in the presence of TTX or from synaptosomes. In many other cases, the suggestion that a presynaptic 5-HT heteroreceptor is operative on an axon terminal (neuron II) (Fig. 1) has been derived from superfusion experiments without TTX on slices prepared from brain areas that contain axons and axon varicosities, but no cell bodies, of the respective neuron (neuron II). The involvement of a 5-HT heteroreceptor on these varicosities themselves was the simplest and most plausible interpretation, which is generally accepted. In particular, this conclusion could be drawn when identical data, obtained on slices in the presence of TTX or on synaptosomes, were available from other investigations. However, in the absence of TTX, alternatives cannot be excluded. Thus, the axon terminal of neuron II, which is subject to modification of transmitter release may be under control of an interneuron (Fig. 2) between neuron I and neuron II. In fact, the 5-HT heteroreceptor may be located on the axon terminals or the somadendritic area of such an interneuron, whose transmitter GABA finally modifies the release of neuron II transmitter. Furthermore, as an even more complex alternative, a circuit consisting of several interneurons within the brain area under study is conceivable, with the first member of the circuit being the site of action of the 5-HT receptor ligand.
In other experiments on slices in the absence of TTX or in intracerebral in vivo microdialysis experiments in which receptor ligands were injected systemically (section II.B.), the presence of 5-HT heteroreceptors mediating inhibition of neuron II transmitter release on an inhibitory GABAergic interneuron is by far the most reasonable explanation (Fig. 2). This is the case when agonists at receptors increasing cell function, e.g., 5-HT3 receptor agonists, inhibit transmitter release from neuron II. Basically, 5-HT3 receptors occur somadendritically and preferentially on the terminals of neurons (Kidd et al., 1993; Boschert et al., 1994; Sari et al., 1999). Activation of somadendritic 5-HT3 heteroreceptors induces mainly Na+ and K+ flux through the receptor ion channel leading to strong depolarization and generation of action potentials propagating to the axon terminals, where Ca2+ influx occurs via voltage-gated Ca2+ channels (Derkach et al., 1989; Hargreaves et al., 1994). In nerve terminals this sequence of events is not valid. As shown on synaptosomes from rat striatum by means of confocal microscopy imaging, 5-HT3 receptor channels on nerve terminals (potentially different in subunit composition from those in the perikaryon) appear to be exclusively Ca2+-permeable and their activation induces Ca2+ influx into the varicosities in a manner independent of voltage-gated Ca2+ channels (Rondé and Nichols, 1998). Irrespective of the somadendritic or presynaptic location of the 5-HT3 receptor in the neuron under consideration, administration of a 5-HT3 receptor agonist should lead to an increase, rather than an inhibition, of the release of its transmitter. Therefore, the inhibition of transmitter release from neuron II by 5-HT3 receptor activation can be most convincingly explained by 5-HT3 receptors located on inhibitory GABAergic interneurons (Fig. 2). They invert the excitatory or facilitatory effect of 5-HT3 receptor stimulation to an inhibition of transmitter release from neuron II. Typical examples for such a constellation are the 5-HT3 agonist-induced inhibitions of acetylcholine release in the rat entorhinal cortex (Barnes et al., 1989) and in the guinea pig neocortex (Bianchi et al., 1990). Most plausibly, the 5-HT3 receptor should not be located on the cholinergic axon terminals but rather on varicosities or cell bodies of inhibitory interneurons, probably GABAergic neurons (Fig. 2) (section III.E.), which are abundant throughout the brain. Recent investigations of the synaptic bouton preparation on mechanically isolated GABAergic interneurons (Katsurabayashi et al., 2003) and on single GABAergic neurons in microcultures (Dorostkar and Boehm, 2005, 2007) unequivocally proved the expression and operation not only of 5-HT3 but also of 5-HT1A heteroreceptors on GABAergic axon terminals.
In most of the in vivo investigations on presynaptic 5-HT heteroreceptors, the intracerebral microdialysis technique has been applied. In a major part of microdialysis studies, 5-HT receptor agonists and/or antagonists were directly administered via the dialysis probe to a certain brain area containing axon terminals (but no cell bodies) of the nonserotoninergic neuron under investigation (neuron II) (Figs. 1 and 2). When in such experiments 5-HT receptor ligands caused a modulation of transmitter release the authors postulated, as a rule, that the 5-HT heteroreceptors involved are located presynaptically, and this suggestion represents a probable and plausible interpretation of the data (Fig. 1). Nevertheless, the same restrictions as in the in vitro experiments on slices in the absence of TTX are valid in such in vivo studies, i.e., the possibility of an involvement of a local circuit of interneuron(s) (Fig. 2) cannot be excluded. However, if the results of such in vivo experiments are identical to those obtained on slices in the presence of TTX or on synaptosomes, the in vivo operation of a presynaptic 5-HT heteroreceptor becomes likely.
Data obtained in in vivo microdialysis (or push-pull perfusion or cup superfusion) studies, in which systemically applied 5-HT receptor ligands modulate nonserotonin transmitter release, raise a particular problem: they can only be interpreted in terms of a presynaptic location of a 5-HT heteroreceptor if they do not substantially differ from the results of in vitro experiments and/or at least from data of in vivo microdialysis investigations in which the 5-HT receptor ligands are administered locally. In other words, suggesting a presynaptic location of a 5-HT heteroreceptor based exclusively on microdialysis investigations in which 5-HT receptor ligands are systemically administered is questionable. For a more reliable interpretation the data must be related to those of appropriate in vitro and local microdialysis studies.
Comparison of local application and systemic injection of drugs in in vivo microdialysis studies provides hints at whether or not the drugs pass the blood-brain barrier. Furthermore, such a comparison makes it possible to get an idea about the overall importance of the presynaptic 5-HT heteroreceptor ligands applied systemically in the control of transmitter release.
|
III. Modulation of Cholinergic Neurotransmission |
|---|
|
TopPreviousNext |
|---|
Pharmacological in vitro experiments examining the operation of 5-HT heteroreceptors on cholinergic neurons have been performed predominantly in the rat brain and to a minor extent in the guinea pig or human brain. According to the presence of serotoninergic as well as cholinergic innervation in certain brain areas, investigations of serotoninergic modulation of ACh release have been performed in the cortex, hippocampus, or striatum. The cerebral cholinergic system plays a major role for memory and cognition, and an increase in ACh concentration in the synaptic cleft represents a therapeutic option in dementia associated with Alzheimer's disease. The role of cholinergic-serotoninergic interactions in cognition has been extensively reviewed (Cassel and Jeltsch, 1995; Steckler and Sahgal, 1995; Buhot, 1997; Feuerstein and Seeger, 1997; Ruotsalainen et al., 1998; Buhot et al., 2000). The interneuronal, somadendritic, and/or presynaptic location of any 5-HT receptor type mediating an increase in ACh release (and, hence, representing a putative target for pharmacotherapy of Alzheimer's dementia) will be discussed with particular care.
It will become evident in the following sections that a clear-cut statement concerning the location of the various 5-HT receptor types is often very difficult. For the 5-HT1 receptor family, the discussion of location and function is particularly complex for four reasons: 1) there is a functional antagonism between the 5-HT1A and 5-HT1B receptors; 2) the receptor under consideration may be located both presynaptically and somadendritically; 3) it may be located on noncholinergic interneurons; and 4) the nomenclature of 5-HT1B/1D receptors has been changed in the meantime. As will be outlined in some detail, presynaptic 5-HT1B/1D receptors played a crucial role in the identification of the main common criterion underlying a rational and unifying classification. A possible location of 5-HT receptors modulating ACh release on noncholinergic interneurons also is involved in the context of 5-HT2 and 5-HT3 receptors.
B. 5-Hydroxytryptamine 1A Receptors
1. Function, Location, and Classification.
In vivo experiments in which 5-HT1A receptor ligands were administered systemically revealed that 8-OH-DPAT and/or other 5-HT1A receptor agonists such as buspirone and ipsapirone [2-[4-[4-(2-pyrimidinyl)-1-piperazinyl-]butyl]-1,2-benzisothiazol-3(2H)-one 1,1-dioxide (5-HT1A agonist)] increase(s) ACh release in the hippocampus and cortex of guinea pigs and rats [using epidurally implanted cups (Bianchi et al., 1990; Siniscalchi et al., 1991) or microdialysis (Wilkinson et al., 1994; Ichikawa et al., 2002)]; this effect was sensitive to blockade by 5-HT1A receptor antagonists. Analogous results were obtained in the spinal cord of anesthetized rats in which a microdialysis probe was placed dorsally at approximately the C5 level (Kommalage and Hoglund, 2005). The selective 5-HT1A receptor antagonist WAY 100635, given alone, inhibited ACh release in the rat frontal cortex, probably by counteracting the release-increasing effect of endogenous 5-HT (Consolo et al., 1996). With respect to the location of the 5-HT1A receptor involved, different possibilities based on contradictory findings have been considered.
By using immunohistochemical double staining it was demonstrated (Kia et al., 1996a,b,c) that 5-HT1A receptors occur on cholinergic cell bodies in the septum and project to the hippocampus and neocortical areas. Concerning the functional significance of septal 5-HT1A receptors, it was shown in slices using intracellular recording techniques that 5-HT1A receptors in the septum mediate hyperpolarization (Van den Hooff and Galvan, 1992). Analogously, using intracellular current clamp recording on immunohistochemically identified cholinergic nucleus basalis neurons (projecting to the neocortex) in guinea pig basal forebrain slices, 8-OH-DPAT and 5-HT caused a hyperpolarization of these neurons, suggesting that the drugs inhibit tonic firing via somadendritic 5-HT1A receptors (Khateb et al., 1993). These results do not conform to the increase in ACh release found in the in vivo experiments. An explanation may be that inhibitory 5-HT1A receptors are present on the cholinergic neurons of the septum and nucleus basalis, i.e., anatomically and functionally homogenous structures crucially involved in learning and memory (Buhot et al., 2000), whereas in the in vivo experiments in hippocampus and neocortex the released ACh originates from less focused neuronal territories in which 5-HT induces via 5-HT1A receptors either directly or indirectly a facilitation of ACh release. The possibility of a direct facilitation of ACh release via presynaptic 5-HT1A heteroreceptors on the axon terminals of cholinergic neurons was ruled out by the finding that 8-OH-DPAT did not influence stimulation-evoked [3H]ACh release from guinea pig hippocampal synaptosomes (Harel-Dupas et al., 1991), guinea pig cortical slices (Bianchi et al., 1990), and rat hippocampal slices (Maura et al., 1989); analogously, a rather preliminary in vivo study revealed that 8-OH-DPAT applied directly to the guinea pig hippocampus via the dialysis probe did not modify ACh release in this brain region (Wilkinson et al., 1994).
In the context of the facilitation of ACh release observed in vivo, it should be kept in mind that 5-HT1A receptors are, as a rule, coupled to Gi/o proteins (Hamon, 1997) and mediate inhibitory effects. Coupling to other signal transduction pathways has been shown under optimized in vitro conditions for recombinant 5-HT1A receptors in various cell lines but not in isolated tissues or in vivo (Raymond et al., 1999). Hence, non-Gi/o protein coupling is a very improbable alternative to explain the facilitatory effect of 5-HT1A receptor agonists on ACh release in vivo. Therefore, it is likely that the 5-HT1A receptor is located on a noncholinergic neuron directly or indirectly innervating the ACh neuron (Izumi et al., 1994; Lüttgen et al., 2005).
One explanation of the 5-HT1A receptor-mediated facilitatory effect on ACh release reported in the literature is based on the assumption that serotoninergic neurons exert an indirect inhibitory tone on cholinergic neurons (Bianchi et al., 1990). In this inhibition a 5-HT3 receptor-mediated stimulation of inhibitory GABAergic interneurons (section VI.) is probably involved. In detail, the authors suggested that 8-OH-DPAT, by activating the inhibitory somadendritic autoreceptors on the serotoninergic neurons, inhibits the activity of the serotoninergic neurons; the inhibition may be assumed to lead to a decreased stimulation of the inhibitory GABAergic interneurons, which, in turn, results in a disinhibition of the cholinergic neurons (Fig. 2).
An alternative explanation of the facilitatory effect of systemically applied 5-HT1A receptor agonists is founded on additional in vivo experiments in which 8-OH-DPAT and other 5-HT1A receptor ligands were administered to the rat hippocampus via the dialysis probe (Izumi et al., 1994). This administration mimicked the effect of systemically injected 8-OH-DPAT, i.e., increased ACh release in a manner sensitive to blockade by the (partial) 5-HT1A receptor antagonist NAN-190. These findings are compatible with the suggestion that the inhibitory 5-HT1A receptor is located on inhibitory hippocampal interneurons, most probably GABAergic interneurons that are densely innervated by serotoninergic afferents (Freund et al., 1990; Halasy et al., 1992) and whose synaptic boutons are endowed with presynaptic heteroreceptors (Katsurabayashi et al., 2003) (section VI.): the 5-HT1A receptor-mediated inhibition of the release of the inhibitory transmitter GABA may be assumed to result in an increase of ACh release from GABAergically innervated hippocampal (and presumably also neocortical) cholinergic neurons. Thus, a presynaptic 5-HT heteroreceptor, albeit on a noncholinergic neuron, appears to be involved in the increase in ACh release in vivo.
2. Concluding Summary Statement. One or both of the suggested sites of action of systemically administered 5-HT1A receptor agonists, i.e., the inhibitory somadendritic 5-HT1A autoreceptor and/or the inhibitory 5-HT1A receptor on GABAergic boutons, is/are probably involved in the indirect facilitation of ACh release in vivo (Table 11). Morphological and electrophysiological investigations revealed that in the septum and nucleus basalis the cholinergic neurons are directly inhibited by 5-HT1A receptors on their cell bodies and dendrites.
C. 5-Hydroxytryptamine 1B and 5-Hydroxytryptamine 1D Receptors
1. Function and Location.
Early evidence that serotoninergic input exerts an inhibition of striatal cholinergic neuronal activity was provided by a study on rat striatal slices in which ouabain-induced ACh release was measured (Vizi et al., 1981). The slices were prepared from rats whose serotoninergic neurons had been destroyed either by electrolytic lesioning of raphe nuclei or chemically by pretreatment with p-chlorophenylalanine or 5,7-dihydroxytryptamine (5,7-DHT), and in these slices ACh release was increased. Conversely, the 5-HT releaser d-fenfluramine reduced ACh release in striatal slices from animals with intact serotoninergic nerves, an effect that did not occur in slices from rats that had undergone chemical lesioning by p-chlorophenylalanine or 5,7-DHT. These findings indicate that endogenous 5-HT released by d-fenfluramine from intact striatal serotoninergic axon terminals exerts an inhibitory effect on striatal cholinergic neurons. The data suggest that under physiological conditions striatal ACh release is tonically inhibited by 5-HT released from serotoninergic raphe-striatal neurons. The authors postulate the involvement of 5-HT heteroreceptors on the cholinergic nerves and their partial location on cholinergic axon terminals.
Recent studies on hippocampal slices from rats whose serotoninergic neurons were lesioned by intracerebroventricular injection of 5,7-DHT confirmed the data of Vizi et al. (1981) in that increased ACh release was found (Birthelmer et al., 2002, 2003a). This could be ascribed to the attenuated activation of the inhibitory presynaptic 5-HT heteroreceptors on cholinergic projection terminals in the hippocampus. Clear evidence for 5-HT1B heteroreceptors on cholinergic terminals in the hippocampus came from experiments on 5-HT1B-receptor knockout mice in which the 5-HT1B agonist CP-93129 did not inhibit ACh release, whereas the 5-HT receptor agonist reduced ACh release in wild-type mice (Rutz et al., 2006).
Table 2 summarizes investigations in which inhibitory effects of 5-HT receptor agonists and/or 5-HT-releasing compounds on ACh release were found. These studies clearly allow the conclusion that the receptors involved in the inhibition are located on cholinergic axon terminals because 5-HT itself and/or 5-HT receptor agonists inhibit the K+-evoked [3H]ACh release from synaptosomes prepared from rat (Maura and Raiteri, 1986; Bolanos and Fillion, 1989; Bolaños-Jiménez et al., 1994) and guinea pig hippocampus (Harel-Dupas et al., 1991). Hence, these receptors on cholinergic axon terminals may be denoted as presynaptic 5-HT heteroreceptors (Fig. 1). In analogy to these results, it is probable that the same type of presynaptic 5-HT heteroreceptor on cholinergic nerves is also involved in the inhibition of electrically evoked [3H]ACh release in slices of the rat hippocampus (Maura et al., 1989; Cassel et al., 1995; Vizi and Kiss, 1998), guinea pig caudate nucleus (Bianchi et al., 1989), and human neocortex (Feuerstein and Seeger, 1997) as well as K+-induced [3H]ACh release in rat striatal slices (Gillet et al., 1985; Jackson et al., 1988). A decrease in electrically evoked [3H]ACh release by the 5-HT receptor agonist CP-93129 (in the absence of TTX) was also observed (5–9 months after grafting) in hippocampal slices from rats that had undergone electrolytic (Cassel et al., 1995) or aspirative (Suhr et al., 1999) fimbria-fornix lesions and 2 weeks later had received intrahippocampal suspension grafts of fetal septal tissue. The authors suggested on the basis of their results that inhibitory presynaptic heteroreceptors are also operative in grafted cholinergic neurons contained in hippocampal slices. Furthermore, in hippocampal slices from aged (25–27 months) rats with various levels of memory impairment shown in a spatial referencememory task, there was no change in CP-93129-induced inhibition of [3H]ACh release (Birthelmer et al., 2003b); in these experiments basically conditions identical to those in slices from grafted hippocampus were applied. The results indicate that aging does not alter the function of presynaptic 5-HT heteroreceptors on cholinergic nerves. Obviously, the presynaptic 5-HT heteroreceptor mediating inhibition of ACh release in the rat hippocampus is also operative in vivo because in a microdialysis study (Izumi et al., 1994) such an effect was observed in response to CGS 12066B and, after blockade of 5-HT1A receptors by NAN-190, also to 5-HT and clomipramine (Table 2). There has been some controversy concerning the 5-HT receptor type that mediates the inhibitory effects of 5-HT in the rat hippocampus because this inhibition requires functional elimination of substance P interneurons endowed with 5-HT2A receptors. It has been suggested that 5-HT1B heteroreceptors on cholinergic terminals reduce, whereas 5-HT2 receptors on substance P interneurons indirectly enhance, ACh release, indicating that in addition to GABA interneurons, further interneurons may be involved (Feuerstein et al., 1996).
2. Classification and Development of Nomenclature.
a. 5-Hydroxytryptamine 1B/1D ligands.
A major problem in the classification of a given 5-HT receptor as 5-HT1B/1D is the shortage of selective agonists and antagonists that discriminate these receptors from the other 5-HT receptors. This problem was particularly true at the time when most of the experiments designed to classify the inhibitory presynaptic 5-HT heteroreceptors were performed, i.e., roughly in the 10 years after 1985. Among the agonists applied for this purpose at least a relative preference for 5-HT1B/1D receptors can be ascribed to CGS 12066B, RU 24969 and 5-CT. Only a few antagonists that were able to counteract the activating effect of 5-HT receptor agonists at the 5-HT1B and 5-HT1D receptors were available; however, drugs such as methiothepine and metergoline are nonselective in that they also block other 5-HT receptors.
The studies performed in that period revealed different pharmacological properties of the presynaptic 5-HT heteroreceptors on cholinergic nerve terminals in the rat brain on the one hand and the guinea pig and human brain on the other. The receptor expressed in the rat brain was named 5-HT1B and that in the brain of guinea pig and humans was named 5-HT1D. Subsequent investigations comprising molecular biological techniques revealed that 5-HT1B and 5-HT1D receptors are different entities coexisting in each of these species. 5-HT1B receptors in the guinea pig, rabbit, and human brain, although encoded by orthologous genes, exhibit pharmacological properties that are different from those of the rat 5-HT1B receptor. New selective 5-HT1B and 5-HT1D ligands finally made it possible and necessary to revise the nomenclature of guinea pig and human presynaptic 5-HT receptors; they also belong to the 5-HT1B subclass. In this context, presynaptic autoreceptors that, in conjunction with 5-HT heteroreceptors, played an additional important role in the development of this classification are relevant.
b. Rat 5-hydroxytryptamine 1B receptor.
The most comprehensive studies in the rat hippocampus (Bolanos and Fillion, 1989; Maura et al., 1989) revealed 1) that 5-HT, RU 24969, and TFMPP inhibited [3H]ACh release, 2) that methiothepine, but not the non-5-HT1B antagonists spiperone and ketanserin [3-[2-[4-(4-fluorobenzoyl)piperidino]ethyl]-2,4-(1H,3H)-quinazolinedione tartrate (5-HT2A antagonist)], counteracted the inhibition, and 3) that methiothepine, but not ketanserin, given alone increased [3H]ACh release probably by preventing endogenous 5-HT from activating the presynaptic 5-HT heteroreceptor. These data were compatible with the conclusion that the latter receptor belongs to the 5-HT1B type.
The inhibitory effect of 5-HT receptor agonists on [3H]ACh release is mimicked in slices and in vivo by 5-HT uptake blockers such as 6-nitroquipazine (Maura et al., 1989), clomipramine (Izumi et al., 1994), and fluoxetine (Jackson et al., 1988) or by the 5-HT releasing drug d-fenfluramine (Maura et al., 1989). Both types of compounds, because of their specific effects on 5-HT inactivation in, and release from, serotoninergic neurons, respectively, increase the concentration of endogenous 5-HT in the biophase of the receptor.
In view of the low number of 5-HT receptor ligands examined in most of the studies of ACh release in the rat hippocampus and striatum summarized in Table 2, each of these studies alone would not be suitable to draw a reliable conclusion concerning the 5-HT receptor type involved. However, the effects produced by each of the few ligands applied in the investigations (Maura and Raiteri, 1986; Bolaños-Jiménez et al., 1994;Izumi et al., 1994) in the rat hippocampus conform to the functional and pharmacological properties of the inhibitory presynaptic 5-HT1B heteroreceptors, thus leading the authors to suggest that their study is dealing with this receptor type, a conclusion that appears to be justified.
In two studies on rat striatal slices (Gillet et al., 1985; Jackson et al., 1988) (Table 2), the ability of methysergide and cinanserin to antagonize the inhibitory effect of endogenous 5-HT (available at increased concentration in the presence of fluoxetine) or of exogenous 5-HT receptor agonists points to the operation of another inhibitory receptor type, putatively in addition to the presynaptic 5-HT1B heteroreceptor. A possible candidate would be a 5-HT2 receptor. Because of this uncertainty, the authors concluded more cautiously that a 5-HT1 receptor or even less specifically that a 5-HT receptor is involved.
In the context of the 5-HT1B character of the inhibitory 5-HT heteroreceptor on cholinergic axon terminals in the rat hippocampus (and putatively also in the rat striatum where it may be coexpressed with a non-5-HT1B receptor), it should be noted that the presynaptic 5-HT autoreceptor in the brain of this species also belongs to the 5-HT1B class defined by radioligand binding in rat brain membranes (Engel et al., 1986). This observation is of particular interest because presynaptic autoreceptors played a significant role in the development of the 5-HT1B/1D receptor nomenclature.
c. Development of guinea pig and human 5-hydroxytryptamine 1B and 5-hydroxytryptamine 1D auto- and heteroreceptor nomenclature.
Radioligand binding experiments on guinea pig cerebral membranes (Waeber et al., 1989) analogous to those in rat brain membranes revealed (despite many similarities) clearly different pharmacological properties of the respective binding sites, leading the authors to the conclusion that a different receptor type named "5-HT1D" is involved, a nomenclature that was premature according to our current knowledge. The same pharmacological characteristics as for the "5-HT1D" binding sites on guinea pig cerebral membranes were established for the presynaptic autoreceptor in the guinea pig brain which, hence, was also classified as "5-HT1D" (Hoyer and Middlemiss, 1989; Limberger et al., 1991). This "old" nomenclature, although revised in the meantime (see below), will be used first (but in quotation marks) to avoid discrepancies with the cited original articles in which it was applied.
Taking the identity of presynaptic 5-HT heteroreceptors on cholinergic nerves and presynaptic 5-HT autoreceptors in the rat brain into account, the questions arose 1) whether the inhibitory presynaptic 5-HT heteroreceptor on cholinergic nerves is a member of the same receptor class as the presynaptic 5-HT autoreceptor in the guinea pig also, i.e., is a "5-HT1D" receptor and 2) which drugs available at that period were suitable to discriminate the guinea pig "5-HT1D" receptor from the rat 5-HT1B receptor.
The latter question was answered by comparing the results in the radioligand binding experiments in rat and guinea pig brain membranes mentioned above. Examples for differences between rat 5-HT1B and guinea pig "5-HT1D" binding sites are the properties of β-adrenoceptor antagonists such as propranolol and cyanopindolol [4-(3-((1,1-dimethylethyl)amino)-2-hydroxypropoxy)-1H-indole-2-carbonitrile (5-HT1A/1
