5-HT_1A and 5-HT_1B receptor agonists and aggression: A pharmacological challenge of the serotonin deficiency hypothesis

Abstract

More than any other brain neurotransmitter system, the indolamine serotonin (5-HT) has been linked to aggression in a wide and diverse range of species, including humans. The nature of this linkage, however, is not simple and it has proven difficult to unravel the precise role of this amine in the predisposition for and execution of aggressive behavior. The dogmatic view that 5-HT inhibits aggression has dominated both pharmacological research strategies to develop specific and effective novel drug treatments that reduce aggressive behavior and the pharmacological mechanistic interpretation of putative serenic drug effects. Our studies on brain serotonin and aggression in feral wild-type rats using the resident–intruder paradigm have challenged this so-called serotonin deficiency hypothesis of aggressive behavior. The well-known fact that certain 5-HT_1A/1B receptor agonists potently and specifically reduce aggressive behavior without motor slowing and sedative effects is only consistent with this hypothesis under the assumption that the agonist mainly acts on the postsynaptic 5-HT_1A/1B receptor sites. However, systemic injections of anti-aggressive doses of 5-HT_1A and _1B agonists robustly decrease brain 5-HT release due to their inhibitory actions at somatodendritic and terminal autoreceptors, respectively. The availability of the novel benzodioxopiperazine compound S-15535, which acts in vivo as a preferential agonist of the somatodendritic 5-HT_1A auto-receptor and as an antagonist (weak partial agonist) at postsynaptic 5-HT_1A receptors, allows for a pharmacological analysis of the exact site of action of this anti-aggressive effect. It was found that, similar to other prototypical full and partial 5-HT_1A and/or 5-HT_1B receptor agonists like repinotan, 8-OHDPAT, ipsapirone, buspirone, alnespirone, eltoprazine, CGS-12066B and CP-93129, also S-15535 very effectively reduced offensive aggressive behavior. Unlike the other ligands, however, a remarkable degree of behavioral specificity was observed after treatment with S-15535, in that the anti-aggressive effects were not accompanied by inhibiting (like other 5-HT_1A receptor agonist with moderate to high efficacy at postsynaptic 5-HT_1A receptors) or enhancing (like agonists with activity at 5-HT_1B receptors and alnespirone) non-aggressive motor behaviors (e.g., social exploration, ambulation, rearing, and grooming) beyond the range of undrugged animals with corresponding levels of aggression. The involvement of 5-HT_1A and/or 5-HT_1B receptors in the anti-aggressive actions of these drugs was convincingly confirmed by showing that the selective 5-HT_1A receptor antagonist WAY-100635 and/or the 5-HT_1B receptor antagonist GR-127935, while inactive when given alone, effectively attenuated/prevented these actions. Furthermore, combined administration of S-15535 with either alnespirone or CGS-42066B elicited a clear additive effect, indicated by a left-ward shift in their dose–effect curves, providing further support for presynaptic sites of action (i.e., inhibitory somatodendritic 5-HT_1A and terminal 5-HT_1B autoreceptors). These findings strongly suggest that the specific anti-aggressive effects of 5-HT_1A and 5-HT_1B receptor agonists are predominantly based on reduction rather than enhancement of 5-HT neurotransmission during the combative social interaction. Apparently, normal display of offensive aggressive behavior is positively related to brief spikes in serotonergic activity, whereas an inverse relationship probably exists between tonic 5-HT activity and abnormal forms of aggression only.

Introduction

A considerable part of our current knowledge on the ethology, neurobiology and, in particular, pharmacology of normal and deviant forms of human aggression is based on experimental laboratory models of aggressive behaviour in animals. Currently, much of this pre-clinical aggression research is conducted in territorial male or maternal female resident rats/mice confronting an intruder conspecific (resident–intruder conflict paradigm). By recording the frequencies, durations, latencies and temporal and sequential patterns of all the observed behavioral acts and postures in the combatants during these confrontations, a detailed quantitative picture (ethogram) of offensive (resident) and defensive (intruder) aggression is obtained (see Olivier and Young, 2004, Miczek et al., 2004, Miczek and de Boer, 2005 for methodological reviews). Using this experimental approach, classical neuroanatomical tracing and electrochemical lesion/stimulation studies have revealed the global neural substrates of aggression (Luiten et al., 1985, Gregg and Siegel, 2001). Several regions of the prefrontal, insular and cingulate cortices, amygdala, septum, medial preoptic area, hypothalamus, periaqueductal gray and their interconnected structures are among the best documented and characterized in this respect. In addition, more recent studies using functional immediate-early gene expression mapping (i.e., c-FOS, p-CREB, zif-268) start to yield a more detailed picture of the individual neurons and their neurochemical identities that become activated within these brain regions during the expression of aggressive behavior (Gammie and Nelson, 2001, Hasen and Gammie, 2005, Kollack-Walker and Newman, 1995, Delville et al., 2000, Halasz et al., 2002, van der Vegt et al., 2003b, Veening et al., 2005). Furthermore, traditional neuropharmacological studies and new molecular genetic approaches (e.g., gene deletion/overexpression/polymorphism) have resulted in an impressive list of molecular substrates (i.e., neurotransmitters/hormones/cytokines and their respective enzymes, receptors and intraneuronal signalling molecules) that, within this neural circuitry, may profoundly influence resident–intruder aggression (see Nelson and Chiavegatto, 2001, Miczek et al., 2002 for excellent reviews). Despite this expansion, most of these molecules appear to act either directly or indirectly on diverse components of the brain serotonin (5-HT) system. Obviously, this classical neurotransmitter system remains the primary molecular determinant of aggression.

For more than 40 years now, this phylogenetically ancient and anatomically very well conserved indolamine system has been postulated to be essential in the control of aggressive and impulsive behavioral traits in many animal species, ranging from invertebrates like fruit flies, crickets and lobsters (e.g., Kravitz and Huber, 2003) to vertebrates like reptiles (Summers et al., 2003, Summers et al., 2005), fish (Overli et al., 1999, Perreault et al., 2003) birds (Ison et al., 1996), mammals (Miczek et al., 2002) and primates including humans (Tuinier et al., 1995, Berman et al., 1997). Therefore, it should not be surprising that among the most promising pharmacotherapeutic approaches to manage violent and excessively aggressive subjects are those that specifically target 5-HT signaling mechanisms (e.g., activating/inhibiting 5-HT synthesis, release, reuptake, degradation and receptor systems).

Numerous studies over the past two decades have very convincingly shown that pharmacological compounds that activate or antagonize 5-HT_1A/IB or 5-HT_2A/C receptor subtypes, respectively, potently suppress the display of aggressive behavior in various animal species ranging from invertebrates, fish, rodents, guinea pigs to primates, including man (Tompkins et al., 1980, Olivier, 2001, Benton et al., 1983, Flannelly et al., 1985, Lindgren and Kantak, 1987, Blanchard et al., 1988, McMillen et al., 1988, Parmigiani et al., 1989, Coccaro et al., 1990, White et al., 1991, Sijbesma et al., 1991, Mos et al., 1992, Mos et al., 1993, Mos et al., 1996, Nikulina, 1991, Sanchez et al., 1993, Sanchez et al., 1996, Sanchez and Hyttel, 1994, Olivier et al., 1989, Olivier et al., 1991, Olivier and Mos, 1992, Olivier et al., 1995, Olivier, 2004, Bell and Hobson, 1994, Bonson et al., 1994, Muehlenkamp et al., 1995, Miczek et al., 1989, Miczek et al., 1994, Miczek et al., 1998, Miczek et al., 2002, Miczek et al., 2004, Cologer-Clifford et al., 1997, De Almeida and Lucion, 1997, De Almeida et al., 2001, De Almeida and Miczek, 2002, Joppa et al., 1997, Lopez-Mendoza et al., 1998, Simon et al., 1998, Fish et al., 1999, de Boer et al., 1999, Ferris et al., 1999, de Boer et al., 2000, van der Vegt et al., 2001, Buitelaar et al., 2001, Rilke et al., 2001, Sperry et al., 2003, Knyshevski et al., 2005). However, similar to the robust anti-aggressive effects of currently employed pharmacotherapeutic treatments (e.g., anti-psychotics, beta-adrenergic blockers, steroid-derivatives and benzodiazepines), virtually all of these 5-HT_1A/1B receptor agonist and 5-HT_2A/C receptor antagonist drugs lower aggressive displays in dose ranges that also severely induce sedation, motor-inactivity, or stereotypies, e.g. rendering them behaviorally nonspecific. This problem of the behavioral specificity (that is, reducing aggressive behavior without concurrently compromising other non-aggressive elements like, sleep, appetite, body-care, sensory and motor activities beyond the normative range) as well as the exact neurobiological mechanism (specific receptor site and location) by which these agents achieve their anti-aggressive effect remain an area of active research.

The behavioral specificity issue can fairly easily be addressed by employing a detailed ethological analyses of all the behavioral elements performed in the offensive aggressive resident–intruder test situation and a comparison of the drug-induced behavioral profile with that of a comparable aggressivity level of drug-free animals. Using such an ethopharmacological approach in either rats or mice, it has recently been claimed that only certain specific 5-HT_1A receptor agonists (i.e., alnespirone and S-15535; de Boer et al., 1999, de Boer et al., 2000), a mixed 5-HT_1A/1B receptor agonist (i.e., eltoprazine; Olivier et al., 1995) and several specific 5-HT_1B receptor agonists (i.e., CGS-12066b, CP-94,253, anpirtoline, zolmitriptan, sumatriptan; Bell and Hobson, 1994, Fish et al., 1999, De Almeida et al., 2001, Miczek et al., 2004) exert behavioral specific anti-aggressive effects. In particular, it was claimed that agonists acting on the 5-HT_1B receptors have more selective anti-aggressive effects in mice than those acting on 5-HT_1A receptors (Miczek et al., 2004, Olivier, 2004). In order to confirm and extend this finding in rats, one goal of this study was to assess and compare the anti-aggressive properties of various 5-HT_1A and 5-HT_1B receptor selective agonists, with special attention to their behavioral specificity.

Although reversal of the anti-aggressive effects of 5-HT_1A and _1B receptor agonists by selective 5-HT_1A and _1B receptor antagonists like WAY-100635 and GR 127935 can confirm the exact receptor type(s) of action (Miczek et al., 1998, Lopez-Mendoza et al., 1998, Fish et al., 1999, de Boer et al., 1999, de Boer et al., 2000, De Almeida et al., 2001), the determination of whether this occurs via pre- or post-synaptic sites remains a critical and conflicting issue that is much more difficult to reveal (e.g., McMillen et al., 1988, Sijbesma et al., 1991, Mos et al., 1993, Millan et al., 1997b, De Almeida and Lucion, 1997, Sanchez and Hyttel, 1994). Determination of pre- versus postsynaptic mediation of the anti-aggressive response to 5-HT_1A and _1B receptor agonists is conceptually important for unraveling the putative inhibitory role of 5-HT neurotransmission in the display of aggressive behavior. This somewhat dogmatic serotonin deficiency hypothesis of aggressive behavior is mainly based on one of the most frequently reported findings in biological psychiatry that excessively aggressive, violent suicidal and impulsive personality traits in human subjects are associated with reduced levels of serotonin's metabolic product 5-HIAA (5-hydroxyindoleacetic acid) in their lumbar cerebrospinal fluid (CSF) and blunted autonomic/neuroendocrine responses to serotonergic challenges (see Berman et al., 1997, Kavoussi et al., 1997 for reviews). Unfortunately and mistakenly, this inverse relationship has been extended toward both species-normative and functionally adaptive levels of aggression in animals, as well as to their actual state-like display of aggressive behavior. Hence, pharmacotherapeutic research strategies have been aimed towards enhancing 5-HT neurotransmission to correct or compensate for the putative hypofunction or deficiency. Moreover, the pharmacological mechanism of serenic drug effects is generally explained by changes in 5-HT neuron function that heighten the effectiveness of serotonergic signalling at their postsynaptic receptor targets. However, in particular the powerful reduction of 5-HT release by 5-HT_1A and 5-HT_1B receptor agonists at their anti-aggressive dose ranges suggests that the specific inhibition of aggressive behavior is based on a reduction rather than enhancement of 5-HT neurotransmission activity.

It is undisputed that both 5-HT_1A receptors and 5-HT_1B receptors exist as inhibitory autoreceptors on the serotonergic neurons (5-HT_1A on the soma and dendrites in the raphe nuclei; 5-HT_1B on the axon-terminals), whereas they are largely inhibitory postsynaptic heteroreceptors in the serotonin system's terminal fields, on target neurons in several corticolimbic regions that govern aggression (Pineyro and Blier, 1999). However, because of this double localization, systemically administered 5-HT_1A and 5-HT_1B receptor agonists in the anti-aggressive dose-ranges have dual effects on serotoninergic neurotransmission: by acting at the autoreceptors, they very effectively inhibit 5-HT neurotransmission. In contrast, when acting directly at postsynaptic sites, these agonists mimics the effect of 5-HT released, thereby mimicking enhanced 5-HT signalling. Thus, the net effect of a 5-HT_1A or 5-HT_1B receptor agonist on 5-HT_1A and 5-HT_1B signal transfer in postsynaptic areas represents a composite of decreased 5-HT release, resulting from autoreceptor activation, and a direct stimulation of the postsynaptic receptors.

To date, virtually all studies investigating whether the anti-aggressive actions of 5-HT_1A or 5-HT_1B receptor agonists are preferentially mediated by pre- or postsynaptic receptors have relied on either 5-HT lesion/depletion or local intracranial microinjection techniques. It is reasoned that depletion/degeneration of 5-HT-containing neurons via systemic administration of p-Chlorophenylalanine (PCPA; irreversible inhibitor of the 5-HT synthesizing enzyme tryptophan hydroxylase) or intracerebral injection of the 5-HT neurotoxic agent 5,7-dihydroxytryptamine (5,7-DHT) would concomitantly remove/decrease somatodendritic 5-HT_1A and terminal 5-HT_1B autoreceptors. Hence, if the anti-aggressive efficacy of the 5-HT_1A/1B and agonists is diminished after these neurotoxic treatments, it would indicate that the presynaptic receptors are involved. If the efficacy is not changed or even augmented, it would indicate that postsynaptic sites are involved. However, it must be realized that both experimental approaches have several pitfalls and limitations (see Discussion section for an elaboration of these). Not surprisingly therefore, the results from these type of studies are rather equivocal.

For the 5-HT_1A receptor, a particularly interesting pharmacological research tool became recently available to address this question with the synthesis of the benzodioxane compound S-15535 (4-(benzodioxan-5-yl)1-(indan-2-yl)piperazine), a highly selective 5-HT_1A receptor agonist with low intrinsic activity which behaves in vivo as a competitive antagonist at postsynaptic 5-HT_1A receptors and as an agonist at 5-HT_1A autoreceptors (Millan et al., 1993, Millan et al., 1994a, Millan et al., 1997a, Millan et al., 2004). For example, in line with its agonist action at somatodendritic 5-HT_1A autoreceptors in the dorsal raphe nucleus, S-15535 produced a marked inhibition of firing of 5-HT neurons and a decrease in 5-HT release and turnover in their projection areas. Consistent with its antagonist actions at postsynaptic 5-HT_1A receptors, S-15535 dose-dependently and completely antagonized postsynaptically mediated 5-HT_1A responses like spontaneous tail-flicks, flat-body posture and hypothermia (Millan et al., 1994a, Newman-Tancredi et al., 1999, de Boer et al., 2000). In accordance with such a unique pharmacological profile are the potent anxiolytic properties in the relative absence of the disruptive motor, autonomic/endocrine and amnesic actions provoked by the activation of postsynaptic sites (Millan et al., 1997b, Cervo et al., 2000). Recently we demonstrated that S-15535 also exerted powerful and selective anti-aggressive effects, already strongly suggesting that a preferential inhibition of ascending serotonergic pathways underlies this serenic effect. To further provide evidence that a transient reduction in serotonergic neurotransmission characterizes the anti-aggressive effects of 5-HT_1A and 5-HT_1B receptor agonists, this study also investigates the effects of combined administration of S-15535 with either a full postsynaptic 5-HT_1A-receptor agonist (alnespirone) or a specific 5-HT_1B receptor agonist (CGS-12066B). If (part of) the anti-aggressive effects of alnespirone are postsynaptically mediated, then S-15535 should exert an antagonistic action, whereas an additive effect would be expected in the case of a somatodendritic mechanism of action. Similarly, when (part of) the anti-aggressive effects of CGS-12066B are mediated via activation of postsynaptic 5-HT_1B receptors, then S-15535 by inhibiting endogenous 5-HT signaling from these receptor sites should exert an antagonistic action. However, an additive effect could be expected in case of a terminal autoreceptor site of action.

This paper will summarize a set of experiments aimed at additional experimental evidence for the involvement of the somatodendritic 5-HT_1A autoreceptor and the terminal 5-HT_1B autoreceptor in aggressive behavior, and hence aimed at further challenging the view that serotonin inhibits aggression (i.e., the serotonin deficiency hypothesis of aggression).