Aggression and Sexual Behavior: Overlapping or Distinct Roles of 5-HT_1A and 5-HT_1B Receptors

Abstract

Distinct brain mechanisms for male aggressive and sexual behavior are present in mammalian species, including man. However, recent evidence suggests a strong connection and even overlap in the central nervous system (CNS) circuitry involved in aggressive and sexual behavior. The serotonergic system in the CNS is strongly involved in male aggressive and sexual behavior. In particular, 5-HT1A and 5-HT1B receptors seem to play a critical role in the modulation of these behaviors. The present chapter focuses on the effects of 5-HT1A- and 5-HT1B-receptor ligands in male rodent aggression and sexual behavior. Results indicate that 5-HT1B-heteroreceptors play a critical role in the modulation of male offensive behavior, although a definite role of 5-HT1A-auto- or heteroreceptors cannot be ruled out. 5-HT1A receptors are clearly involved in male sexual behavior, although it has to be yet unraveled whether 5-HT1A-auto- or heteroreceptors are important. Although several key nodes in the complex circuitry of aggression and sexual behavior are known, in particular in the medial hypothalamus, a clear link or connection to these critical structures and the serotonergic key receptors is yet to be determined. This information is urgently needed to detect and develop new selective anti-aggressive (serenic) and pro-sexual drugs for human applications.

1. Introduction

Early studies into the role of various brain structures in aggression focused merely on the hypothalamus already starting in the 1920s by Hess [1]. Electrical stimulation in the hypothalamus induced stimulation-evoked (or –bound) attack behavior, where rat studies [2, 3, 4, 5] created an important framework underlying the functional organization of attack circuitry [6]. Lesion experiments in mice [7] and rats [8, 9] also implicate parts of the hypothalamus in offensive and defensive aggression [9, 10].

Anterior hypothalamic lesions, damaging large parts of the anterior hypothalamus (AH), rostral parts of the ventromedial hypothalamus (VMH), and smaller parts of the caudal preoptic area (PA), indeed, strongly increase defensive behavior toward a male intruder. Mammillary body lesions, damaging large parts of the ventral (vPMV) and dorsal premammillary nucleus (dPMV), caudal parts of the arcuate nucleus, medial mammillary nucleus (mMM), posterior mammillary nucleus (pMM), supramammillary peduncle induced strong increases in offensive aggression [9]. These findings suggested the existence of at least two distinct neural substrates in the hypothalamus normally modulating defensive (anterior medial hypothalamus) and offensive (posterior medial hypothalamus) aspects of intermale aggression. Concomitant studies strongly support the ventral premammillary nucleus as a possible central hub of aggression [11]. Because electrolytical lesions are rather nonspecific, i.e., it is virtually impossible to wipe out, on both sides of the brain, one structure without damaging other structures including neurons and fibers of passage. Alternatively, electrical (micro) stimulation can be used to study the role of the underlying substrate (again neurons and fibers of passage) in certain behaviors, including aggression (and sexual behavior).

Already early in the twentieth century [1], hypothalamic stimulation in cats induced attacks. In the seventies [2, 3], this research extended to rats where electrical stimulation in the (ventrolateral) hypothalamus induced attack behavior, although rather nonspecific in that different subjects (mice (live or dead), rat pups, guinea pigs, and adult rats) were attacked. At the end of the seventies, the groups of Koolhaas in Groningen and Kruk in Leiden extensively investigated that upon electrical stimulation in hypothalamic structures, specific behavioral responses were elicited [4, 5]. The Kruk group [12] described, after extensive and meticulous studies, an “aggressive area,” later named the “Hypothalamic Aggression Area (HAA),” lateral from the ventrolateral lobe of the ventromedial hypothalamus (VMH) into the frontal pole of the VMH and the anterior hypothalamic nucleus (AHN). This area extends medially to the arcuate nucleus through the ventrolateral and medial parts of the VMH (see for a 3D-picture, Fig. 1 in [13]. The HAA largely (or completely) coincides with the area in the hypothalamus that upon lesioning leads to reduced aggression [9]. In female rats, aggression can be evoked in the same (HAA) area as in males [14].

Recent studies applying genetically defined functional manipulations showed that the VMH and in particular the ventrolateral part (vlVMH) and the medial amygdala are critical sites to evoke aggression [15]. The VMH receives direct and indirect input from the medial amygdala and from the bed nucleus of the stria terminalis (principal nucleus:BNSTpr), but also from various other structures [16], such as the ventral premammillary nucleus (vPMN), the lateral septum (LS), and subparaventricular zone (SPZ) [17].

Newman [18] argued already that the neurobiology of aggressive behavior is embedded in a larger and integrated network of various social behaviors, including sexual and parental behavior. This implies that the neural circuitry involved in these behaviors must consist of a number of commonly activated brain areas (sensitive to a variety of shared cues) and separately by brain areas specifically involved in a specific function or a selective role in each behavior, as, e.g., occurs in the vomeronasal circuitry [19].

The putative “dual” or “multiple” involvement of a specific neural area (node) in, e.g., aggression and sexual behavior necessitates testing the effects of manipulations of this node in other behaviors, e.g., exploratory or other behaviors. Veening and coworkers [20] approached this question by studying whether the patterns of brain activation during male rat aggression and sexual behavior are specific for either behavior or show (partly) overlapping patterns. By using Fos-immunoreactivity, it was shown that some brain areas (caudal medial preoptic area and medial amygdala) were commonly activated, whereas other areas (posterodorsal parts of the medial amygdala, rostral preoptic and premammillary hypothalamus) show differences in neural activation. This is in line with the idea that aggressive and sexual behaviors share partly integrated neural pathways, next to more specific “aggressive” and “sexual” brain areas. In general, the medial preoptic nucleus (MPN) and the VMHvl are essential regions for male sexual and aggressive behavior, respectively. Estrogen receptor alpha (ESR1)-expressing cells in the posterior amygdala (PA) are a main source of excitatory input to the hypothalamus and are main mediators for mating and fighting in male mice [21]. PA^Esr1+ neurons to the MPN are activated during sexual behavior and also induce sexual behavior. PA^Esr1+ neurons that project to the VMHvl promote attacks. The PA can be considered a key node in male aggressive and sexual behavior circuitry. Optogenetic activation of VMHvl cells expressing estrogen receptor alpha-progesterone receptors induced attack, whereas pharmacogenetic optogenetic inactivation of the VMHvl inhibited naturally occurring aggression [17], Moreover, the VMHvl is also involved in generating preparatory (learned) behaviors associated with the attacks [22]. Available evidence gives an essential role to the medial hypothalamus in the generation of aggression, the hypothalamic aggression area or circuit. This hypothalamic aggression circuit is embedded, upstream and downstream, in other circuits that modulate the aggression outcome, e.g., the dopaminergic mesolimbic dopamine pathway. Its connection from the ventral tegmental area (VTA) to the nucleus accumbens (NAc) is a key circuit in the rewarding control of aggression [17, 23].

Optogenetic stimulation of the network in the HAA that evokes aggression [15] was also able to evoke mounting. The neurons involved (estrogen-1 (Esr1)-expressing) in the VMHvl evoke mounting upon stimulation with lower frequencies [24]. The VMHvl neurons are sensitive to varying levels of optogenetic stimulation and the (behavioral) outcome ranges from highly prosocial (sexual) to antisocial (aggression). Apparently, one could assume that the VMHvl, and specifically one type of neuron (Esr1⁺), is an overlapping node in male aggression and sexual behavior circuits [25].

For a long time, it was a common belief that male aggressive and sexual behavior shares many of the underlying neurobiological, neurological, pharmacological, physiological, and neuroendocrine mechanisms. This seems, at least partly, true for aggression and sexual behavior. Apparently, such a shared structure (e.g., the VMHvl) mediates multiple social behaviors and processes [26]). Factors such as social experience, behavioral context, hormonal state, spatial and sensory cues probably (co)-influence which behavior is generated at a specific moment and time [27].

2. Serotonin and sexual behavior

Sexual behavior systems operate under rather constant inhibitory control to ascertain that sexual behavior is performed only under appropriate circumstances. Serotonergic neurotransmission is involved in inhibitory and disinhibitory processes regulating proper sexual behavior. 5-HT release, facilitating transmission, is regulated via negative feedback mechanisms, through different presynaptic (5-HT_1A, 5-HT_1B/1D) autoreceptors. Moreover, postsynaptic serotonergic heteroreceptors are also involved in negative feedback on serotonergic cell firing [28]. The 5-HT transporter (5-HTT or SERT) plays an important role in homeostatic modulation of the magnitude, duration, and spatial distribution of signals reaching serotonin receptors [29, 30]. Although 5-HT is not considered a central modulator of sexual behavior, but rather modulatory or facilitating, 5-HT activity plays an important role during sexual behavior, via its machinery of pre- and postsynaptic interactions, thereby critically interfering with GABA-ergic and glutamatergic neurons in various brain areas (prefrontal cortex, hypothalamus, lateral habenula, and dorsal raphe nucleus). Serotonergic fibers are abundant in many areas of the spinal cord implicated in ejaculatory processes [31]. Postsynaptic 5-HT receptors are located at lumbar spinothalamic cells [32], indicative of a role of 5-HT in ejaculation at the level of the spinal cord, descending from supraspinal areas such as the nPGI. These descending 5-HTergic neurons from supraspinal areas innervate spinal cord mechanisms that control bulbospongiosus muscles, which have inhibitory effects on ejaculation [32]. At hypothalamic level, the medial preoptic area (mPOA) is involved in lowering an ejaculatory threshold via inhibition of an inhibitory serotonergic tone exerted by the nPGI [33, 34], removing a brake on ejaculatory processes. The lateral hypothalamic area (LHA) is also involved in ejaculation: lesions affect ejaculation, but not preceding mounts and intromission [35]. Because 5-HT is released in the LHA at the occurrence of ejaculation and infusion of selective serotonin reuptake inhibitors (SSRIs) into this area influences sexual behavior [36], a role of serotonin is clearly implicated. The main sources of 5-HTergic innervation of the forebrain emerge from the dorsal (DRN) and medial (MRN) raphe nuclei. Ascending 5-HTergic fibers are divided into a meso-limbic pathway from the MRN and a meso-striatal pathway derived from the DRN [31, 37, 38]. Although DRN and MRN have (partly) overlapping projections, they do not overlap in the projected structure but go to different subareas [38]. MRN and DRN have reciprocal connections, and both structures express high densities of 5-HT_1A receptors. An unanswered question is whether and how these extremely complex interactions (including those with non-serotonergic structures) interact during sexual behavior [39, 40]. Most research in these areas is performed in males (mostly rodents).

Notwithstanding an extensive role of serotonin in aggression and sexual behavior [29, 41] in the present chapter, we focus on the role of two receptors, 5-HT_1A and 5-HT_1B receptors because they appear as most relevant in interactions between aggressive and sexual behaviors (Figure 1 shows a cartoon of a serotonergic neuron with all 14 different 5-HT receptors). 5-HT_1A receptors are present as somatodendritic autoreceptors on serotonergic neurons that present upon activation as negative feedback on cell firing, thus inhibiting 5-HT release and thereby exerting a broad influence on 5-HTergic tone. 5-HT_1A receptors are also widely distributed in terminal areas of the brain expressed as postsynaptic heteroreceptors in a variety of different brain structures and influence a wide scale of neuropsychopharmacological events [42]. 5-HT_1B receptors and its counterpart 5-HT_1D receptor have a long, complex, and debated history (see Figure 3 in [42]), because of species differences in function and structure. It was finally confirmed that 5-HT_1B and 5-HT_1D receptors represent two different receptor classes, and the 5-HT_1B-receptor (including the rat 5-HT_1B-receptor) plays the most prominent functional role, although the pharmacology of ligands for the human and rodent 5-HT_1B receptors can be quite deviating. Most of animal behavioral data on 5-HT_1B receptor ligands have been gathered in rodents, which makes prediction for human applications sometimes unreliable [42]. 5-HT_1B receptors are present as inhibitory autoreceptor on the presynaptic part of 5-HT neurons (see Figure 1) and as inhibitory heteroreceptor on non-serotonergic neurons [42]. Although it is unclear whether every single 5-HT neuron is equipped with similar autoreceptors, at least for the MRN and DRN it is known that they possess somatodendritically localized autoreceptors and presynaptically localized 5-HT_1B autoreceptors and 5-HT transporters. 5-HT activity has to be terminated, which is effectuated via reuptake of 5-HT by the serotonin transporter (SERT), a complex molecule with 13 transmembrane loops. After this uptake over the cell membrane via the SERT from the synapse, 5-HT is subsequently taken up by the vesicular-monoamine transporter (VMAT2) and stored in the synaptic vesicles for reuse. Another major route to end serotonergic activity is a process whereby 5-HT is taken up by the surrounding glia cells and degraded by the enzyme monoamine-oxidase-A (MAO-A) [43] to its metabolite 5-hydroxyindole acetic acid (5-HIAA). Simultaneously, the released 5-HT activates 5-HT_1B autoreceptors leading to inhibition of further 5-HT release from the vesicles and activates also the somatodendritical 5-HT_1A autoreceptors, leading to inhibition of cell firing [44, 45]. The interplay between these three mechanisms (5-HT reuptake, inhibition of release via activation of 5-HT_1B autoreceptors, and inhibition of cell firing via activation of somatodendritic 5-HT_1A autoreceptors) reduces the activity of the serotonergic neurons after activity, preparing the neuron for a new discharge [44, 45]. Of course, many non-serotonergic inputs are acting on serotonergic cells in the raphé nuclei (a nice schematic overview is shown in Fig. 2 in [46]).

Figure 1.

Cartoon of a serotonergic neuron projecting to two non-serotonergic neurons. Fourteen different serotonergic receptors are located either as presynaptic autoreceptors (5-HT_1A, 5-HT_1B) or as postsynaptic heteroreceptors (all 14 receptors). The 5-HT transporter (SERT) is located at the somadendritic and synaptic part of the serotonergic neuron. See text for further details.

3. 5-HT_1A and 5-HT_1B receptors in aggressive and sexual behavior

The prototypic 5-HT_1A-receptor agonist (±)-8-OH-DPAT was developed in the early eighties and when tested on male rats, showed a remarkable stimulation of male sexual behavior [47]. The effects of the racemic (±)-8-OH-DPAT and its active enantiomer (+)-8-OH-DPAT have been confirmed in many subsequent studies [31]. Veening and Coolen [48] presented a so-called “funnel-model” of sexual behavior in the rat, based on experiments on feeding, sexual and aggressive behavior combined with electrical stimulation in the ventromedial hypothalamus [49]. The “funnel-model” also applies to other behavioral systems, including aggressive behavior. In general, in the initiation phase 1, the animal involved gathers information about the environment (scanning, sniffing, orientation), followed by transition to phase 2 where appetitive behavior becomes prominent (anogenital sniffing, mounting in case of sexual behavior; following, anogenital sniffing in aggression). Transition to phase 3 may follow, which is the consummatory/executive phase. In case of sexual behavior, this includes mounting, intromission, and finally, ejaculation; in aggression, this includes lateral threat, biting, jump attacks, keeping down, and chasing [9].

Both sexual and aggressive behavior in male rats can be described by a “funnel”-like pattern of behavior [49, 50, 51]. By manipulations such as electrical stimulation or lesions in the (ventro)medial hypothalamus, this pattern of behavioral funneling can be interrupted, e.g., electrical stimulation in the VMH in a resident-intruder situation [49] strongly reduces the chance on full aggression, because the stimulation strongly promotes return to phase 1 (scanning and initiation phase). Remarkably, stimulation (either electrically or optogenetically) of the VMHvl in mice can evoke both aggressive and sexual behavior [25, 52, 53, 54]. Extensive studies indicate that intermingled, antagonistic brain circuits for aggressive and sexual behavior are present in the VMHvl [15, 55].

Administration of 5-HT_1A-receptor agonists (e.g., 8-OH-DPAT, flesinoxan, buspirone, ipsapirone, and others [31]) dose-dependently increases the number of ejaculations and reduces the ejaculation latencies during a certain test duration (e.g., 30 min). Moreover, the number of mounts and intromissions during the successive ejaculation series decrease (Figure 2). This whole profile has been described as “pro-sexual” and can be aligned with the funnel-model hypothesis, assuming that 5-HT_1A-receptor activation strongly drives the direction of sexual behavior toward the final consummatory phase, ejaculation.

Figure 2.

Time course of sexual behavior of male rats treated with vehicle (top), (±)-8-OH-DPAT (middle), and eltoprazine (bottom) at behaviorally active dosages. M = mount, I = intromission, PEL = post-ejaculatory interval, ES = ejaculation series, EL = ejaculation latency, ML = mount latency, IL = intromission latency. Numbers (1, 2, 3, …) indicate in which ejaculation series (ES) the behavior parameter is scored.

Recently, we tested some new 5-HT_1A-receptor agonists on male rat sexual behavior [56]. They are so-called “biased” or “functionally selective” high potency 5-HT_1A-receptor agonists, F15599 and F13714, and have distinct pre- and postsynaptic agonistic activity [57]. However, like in aggression (see later), both “biased” agonists had potent pro-sexual activity, comparable to “classic” 5-HT_1A-receptor agonists such as 8-OH-DPAT or flesinoxan (similar activation of 5-HT_1A auto- and heteroreceptors). However, S-15535, primarily considered a 5-HT_1A-autoreceptor agonist and heteroreceptor antagonist, had no pro-sexual activity at all, and also no sexual inhibitory activity either [56]. This strongly suggests that “pro-sexual” activity induced by 5-HT_1A receptor agonists is primarily caused by activation of postsynaptic 5-HT_1A heteroreceptors.

5-HT_1B-receptor agonists inhibit male sexual behavior in the rat [58, 59, 60] and in the mouse [61]. Eltoprazine, a mixed 5-HT_1A/1B receptor (partial) agonist [62], dose-dependently reduced male rat sexual behavior; at no dose tested, pro-sexual effects were seen, indicating that the putative 5-HT_1A receptor activating effects of eltoprazine were “overshadowed” by the 5-HT_1B-receptor agonistic effects and that the behavioral effects were caused by 5-HT_1B-receptor activation. A comparable mixed 5-HT_1A/1B-receptor agonistic profile in other putative 5-HT_1B receptor agonists such as mCPP, TFMPP, RU24969, and anpirtoline, which all have inhibitory sexual effects in male rats, points to the dominance of 5-HT_1B receptors over 5-HT_1A receptors upon concomitant activation. In mice, in contrast to rats, 5-HT_1A-receptor agonists (8-OH-DPAT) have an inhibitory effect in male sexual behavior [61] (Figure 3).

Figure 3.

Effects of a 5-HT_1A-receptor agonist ((±)-8-OH-DPAT) and a 5-HT_1B-receptor agonist (eltoprazine) on male sexual behavior of Wistar rats. The number of ejaculations (top left), mounts (left bottom), and intromissions (right bottom) and the ejaculation latency (top right-in seconds) are shown. 8-OH-DPAT was subcutaneously administered 30 min before testing; eltoprazine orally 60 min before testing. * indicates significant difference (p < 0.05) from vehicle (0 mg/kg). In the figures of ejaculation latency and number of mounts and intromission, at higher doses of eltoprazine, no data are available because of absence of sexual behavior. Data are derived from [63] and Olivier-unpublished 1991.

In our studies on 5-HT_1A-receptor knockout mice [64, 65, 66], we tested three strains of mice (the background strains used to produce the gene knockouts; 129Sv/Ev, C57Bl/6, and Swiss Webster) in male sexual behavior. Figure 4 shows the data on number of mounts, intromission, and ejaculations and intromission and mount latencies during male/estrus female tests of 1500 s duration (25 min). In 129Sv/Ev and C57Bl/6 strains, wild-type (WT) mice had significant higher sexual behavior levels than the respective 5-HT_1A-receptor knockout mice. Swiss Webster mice hardly showed any sexual behaviors, neither in WT, nor in KO animals, making conclusions impossible.

Figure 4.

Sexual behavior parameters of 5-HT_1A-receptor knockout (5-HT_1AKO) and wild-type (WT) mice of three different strains. Latencies are expressed in seconds. Statistics: Repeated measures analysis with two time points. * P < 0.05; ** P < 0.005; # P = 0.07.

5-HT_1B-receptor knockout mice (in 129/SV-ter strain) have a lower baseline of sexual behavior than the corresponding wild-type mice [61]. TFMPP, a 5-HT_1B-receptor agonist had no behavioral effects in 5-HT_1B receptor knockout mice, whereas it dose-dependently decreased male sexual behavior in wild-type mice. Intriguingly, 8-OH-DPAT also dose-dependently decreased male sexual behavior in WT and had, at these doses, no effect in the KO mice. In another mouse strain (NMRI), 8-OH-DPAT had also inhibitory effects on male mouse sexual behavior [67].

There appears a clear species difference between mice and rats regarding 5-HT_1A-receptor modulation of male sexual behavior. In contrast, such a species difference is not present in 5-HT_1B-receptor modulation. 5-HT_1B-receptor agonists inhibit both male aggression and sexual behavior in mice and rats. Selective 5-HT_1A-receptor antagonists (e.g., WAY100,635) have no intrinsic behavioral effects in either sexual or aggressive behavior, either in mice or rats [68, 69, 70, 71, 72, 73, 74, 75, 76]. No studies have been published on effects of 5-HT_1A-receptor antagonists on male aggression or sexual behavior in 5-HT_1A or 5-HT_1B-receptor knockout mice.

There are limited data on chronic treatment with 5-HT_1A- or 5-HT_1B-receptor agonists on male sexual behavior. Flesinoxan, a classic 5-HT_1A-receptor agonist, was given twice daily for 14 days at 2.5 mg/kg, IP. Animals were tested acutely, subchronically (after 7 days) and chronically (after 14 days) on sexual behavior against an estrus female. Acutely, flesinoxan had pro-sexual effects, but no effects were observed after chronic administration, suggesting some tolerance [77]. The effects of a selective 5-HT_1B-receptor agonist CP-94253 (injected subcutaneously four times daily with 5 mg/kg) were also tested acutely and after 7 and 14 days. CP-94253 inhibited sexual behavior at all time points, showing that it did not induce tolerance [77].

Removing (gene knockout) 5-HT_1A receptors from all neurons normally bearing them (serotonergic somatodendritic autoreceptors) and non-serotonergic neurons with 5-HT_1A heteroreceptors [29] has behavioral consequences for male sexual behavior but not for male aggressive behavior. Removing 5-HT_1B receptors from serotonergic synapses (inhibitory autoreceptors) and inhibitory postsynaptic 5-HT_1B heteroreceptors) has contrasting effects on male sexual and aggressive behavior: male 5-HT_1B receptor knockout mice have reduced sexual behavior, whereas male aggressive behavior is enhanced [78, 79].

Development of a tissue-specific and temporally conditional 5-HT_1B-receptor mouse model [80] brought more insight. It was shown that aggressive behavior is mediated by developmental expression of 5-HT_1B heteroreceptors. Whole-life, whole-brain elimination of 5-HT_1B receptors led to enhanced aggression, like present in constitutive knockout mice [78, 79]. Rescue of 5-HT_1B-receptor expression in early postnatal development, but not in adulthood, ameliorated aggression. It was shown that forebrain 5-HT_1B heteroreceptors mediated this aggression phenotype, while reduction of 5-HT_1B autoreceptors had no effect on aggression. Apparently, a developmental sensitive period exists, during which the presence of serotonin affects the development of adult aggression.

4. Are effects of 5-HT_1A-receptor agonists and 5-HT_1B-receptor agonists on male aggression and sexual behavior mediated by presynaptic (autoreceptor) or postsynaptic (heteroreceptor) serotonin receptors?

The big question is whether specific effects induced by activating very heterogeneous 5-HT_{1 (A or B)} receptors on very specific behavioral systems (aggression and sexual behavior) that are functioning via specific and localized neural circuitry in the brain can be influenced via the extremely nonspecific influence of autoreceptors on the serotonergic cell bodies (5-HT_1A autoreceptors) or the serotonergic synaptic endings (5-HT_1B autoreceptors). Remarkable is at least that selective 5-HT_1A-receptor antagonists, blocking somatodendritic autoreceptors and basically leading to enhanced 5-HT release in serotonergic synapses, do not induce behavioral effects (at least in aggression and sexual behavior). Of course, the enhanced 5-HT levels are also not able to stimulate 5-HT_1A heteroreceptors because 5-HT_1A receptor antagonists block these too, but other 5-HT receptors are not blocked and could be instrumental in emerging behaviors. This apparently does not happen: 5-HT_1A receptor antagonists are generally intrinsically silent, i.e., they do not exert intrinsic behavioral effects [51, 71, 74]. Whether 5-HT_1B autoreceptors plus serotonin transporters completely compensate for the effects of blocking 5-HT_1A receptors on 5-HT release is largely unknown, but seems less likely. Although certainly not conclusive, we postulate that the pro-sexual effects of 5-HT_1A-receptor agonists on male sexual behavior are mediated via postsynaptic 5-HT_1A heteroreceptors. Abundant presence of 5-HT_1A receptors in areas containing (parts of) neural circuitry involved in all aspects of sexual behavior (e.g., the hypothalamic circuitries [52, 55]) makes this a likely hypothesis, although microinjection of selective 5-HT_1A receptor ligands in nodes of these circuits is badly needed.

In aggression, the role of 5-HT_1A receptors is also not evident. Although 5-HT_1A-receptor agonists have strong anti-aggressive effects in various aggression models in rodents (intermale aggression, resident-intruder aggression, colony aggression, isolation-induced aggression, and others [62, 81], these anti-aggressive effects often coincide with associated nonspecific behavioral effects such as sedation or motor retardation [29, 71, 76]. The selective but low efficacy 5-HT_1A-receptor agonist S-15535, acting preferentially as a 5-HT_1A-receptor autoreceptor agonist and as a (partial) 5-HT_1A-heteroreceptor antagonist, rather selectively decreased aggressive behavior [76], suggesting that the “classical” 5-HT_1A-receptor agonists (that activate auto- and heteroreceptors) induce the “nonspecific” anti-aggressive effects via heteroreceptor activation.

The emergence of so-called “biased” or “functionally selective” 5-HT_1A-receptor agonists yielded the possibility to study selectively presynaptic 5-HT_1A-autoreceptors versus postsynaptic 5-HT_1A-heteroreceptors. F15599 is an extremely effective 5-HT_1A-heteroreceptor agonist, with relatively low activity at autoreceptors [57, 82]. F13714 is also an effective 5-HT_1A-receptor agonist, but primarily activates 5-HT_1A-autoreceptors [57, 82]. Both “biased” agonists have anti-aggressive effects in extremely aggressive (violent) semi-wild rats [72]: no difference in their anti-aggressive profile was found, making conclusions about specific roles of pre- versus postsynaptic 5-HT_1A receptors in aggression more complex.

Although all 5-HT_1A-receptor agonists upon acute administration seem to inhibit aggressive behavior in mice and rats, classic 5-HT_1A-receptor agonists such as 8-OH-DPAT and flesinoxan do not inhibit aggression induced by electrical stimulation in the hypothalamic attack area (HAA) in male rats [6, 12, 81, 83]. It is remarkable and unexpected that direct activation of “aggression neurons or circuitry” in the HAA (including the VMHvl) cannot be inhibited by activation of 5-HT_1A receptors in the brain. This sharply contrasts by 5-HT_1B-receptor activation (e.g., by eltoprazine, fluprazine, or TFMPP) that dose-dependently decreases attacks (measured by enhanced stimulation thresholds), but does not influence locomotion thresholds (or even decrease them) and also dose-dependently reduces teeth chattering, an associated (autonomic) aggressive element [6, 12, 81, 83].

5-HT_1B-receptor agonists inhibit offensive aggression in mice, rats, and other species (e.g., monkeys and pigs) [29, 62, 84]. Other groups have confirmed that activation of 5-HT_1B receptors leads to reduction of aggression [71, 85]. Support for an important role of postsynaptic 5-HT_1B receptors has been found by the Miczek group [86, 87] and several other sources [88]. Overwhelming evidence suggests that postsynaptic (heteroreceptor) 5-HT_1B receptors are involved in the mediation of specific anti-offensive aggression (serenic) activity [29]. Considerable efforts still need to be made to unravel the neural localization of these postsynaptic 5-HT_1B receptors, because several conflicting data exist.

A weak and underreported aspect of aggression (and sexual behavior) research is that studies are almost only performed after acute administration. No chronic aggression studies with 5-HT_1A-receptor agonists have been performed as far as we are aware. For 5-HT_1B-receptor agonists, some chronic aggression studies in mice and rats were performed. Fluprazine, an early serenic [90, 91], was tested in wild house mice that were selected for a high level of aggression, measured by the attack latency when confronted with a male opponent [89]. Sixteen wild male mice of the SAL-line were selected for Short Attack Latencies (<100 s) and were trained in three successive 10-min trials to reach a stable short attack latency (Figure 5A-pre value). In the fourth trial, eight mice received saline (IP) and eight mice received fluprazine (20 mg/kg IP, 30 min before testing). Figure 5A shows the strong anti-aggressive activity of acutely administered fluprazine. Chronic administration was performed using osmotic minipumps. Saline had no effects on the latency time, whereas after 7 days of 400 mg/kg/day via minipump administration, fluprazine had strong anti-aggressive activity (Figure 5B). Although the chronic effects of fluprazine seemed diminished compared with acute dosing, it was unknown whether a dose of 400 mg/kg/day led to comparable plasma levels of fluprazine than after acute administration, although some tolerance (desensitization of 5-HT_1B receptors) might be possible. However, this seems unlikely seen the results of a chronic study with eltoprazine in male Tryon Maze Dull (TMD-S3) rats, a strain with a high level of residential aggression [92]. After initial training, male rats were used for the resident-intruder test during a 4-week treatment period. Eltoprazine (0, 1, and 3 mg/kg PO) was given 60 min before a 10-min aggression test. Acutely, eltoprazine reduced offensive aggression without any effects on social and nonsocial behaviors. Subsequently, eltoprazine or vehicles were daily administered for 4 weeks, and a resident-intruder test was performed once weekly. The anti-aggressive effects of eltoprazine remained stable over the 4-week period, whereas exploration was enhanced, but no adverse effects were found. After a washout of 1-week aggressive behavior returned to baseline. These data showed no tolerance for the anti-aggressive effects of eltoprazine [93]. In a comparable study using osmotic minipumps for 7 days with 20-mg/kg/day eltoprazine, also no evidence was found for tolerance to the anti-aggressive activity confirming the specificity of the effects [94].

Figure 5.

Effects of fluprazine on attack latencies (sec) of wild house mice selected for aggressive behavior. Panel A shows the acute effects of vehicle and fluprazine 20 mg/kg IP. Panel B shows the effects after 7 days treatment of chronic vehicle or chronic fluprazine (200 mg/kg SC) administered via Alzet® osmotic minipumps [89].

5. Overview of serotonin, sexual behavior, and aggression

Table 1 summarizes effects of 5-HT_1A and 5-HT_1B-receptor agonists on male sexual and aggressive behavior in mice and rats. 5-HT_1A-receptor agonists enhance male sexual behavior in the rat, but decrease it in the mouse. In contrast, 5-HT_1A-receptor agonists decrease male aggression in most offensive aggression models in mice and rats, although not in a very essential model, HAA stimulation in rats and behavioral effects are often not very specific, and induced side effects such as sedation or sensoric-motor disturbances might be (co)-causative in the reduction of offensive behavior.

Table 1.

Summary of effects of 5-HT receptor ligands on male aggressive and sexual behavior in mice and rats.

 = increase;

 = decrease;

 = no effect; nt = not tested.

In contrast, 5-HT_1B-receptor agonists show highly specific anti-aggressive effects in all offensive aggression models and also reduce male sexual behavior. Neither 5-HT_1A-receptor antagonists, nor 5-HT_1B-receptor antagonists exert any behavioral effects in either male aggression or male sexual behavior models. Chronic administration of 5-HT_1A-receptor agonists seems to induce tolerance, whereas 5-HT_1B-receptor agonists do not.

These profiles favor 5-HT_1B-receptor agonists over 5-HT_1A-receptor agonists with regard to anti-aggressive activity, whereas 5-HT_1A-receptor agonists may have a role in pro-sexual effects that may be useful in certain male human sexual dysfunctions, e.g., delayed ejaculation.

In the following part, the history of the development of specific anti-aggressive (offensive) drugs is depicted (Box 1).

Box 1.

Serenics: Drugs with specific anti-aggressive activity.

The discovery and development of drugs, specifically aimed at reduction of pathological aggression and destructive behavior in psychiatric patients, were started halfway the seventies of last century by Philips-Duphar in the Netherlands. It was already at that time clear that pathological destructive behavior, sometimes named “aggressive,” “violent,” “agitated,” or “dysfunctional,” is widely present in psychiatric and neurological disorders and cannot, even up to this time, adequately be treated with psychotropic drugs. A striking variety of drugs were and are used in patients with these severely troubled behavioral disturbances, including neuroleptics or hypnotics, mainly used for their sedative properties, benzodiazepines, lithium, beta-blockers antidepressants and anticonvulsants [95].

In the mid-seventies, Philips-Duphar started a search for specific anti-aggressive drugs. At that time, the pre-molecular era, there was no clue for which target to search, and consequently, the quest for anti-aggressive drugs was steered by animal aggression models and tests [96]. One of the authors (BO) was hired by Philips-Duphar because of their expertise and background in aggression models and brain mechanisms involved in aggression [29]. Throughout the sixties and seventies, pharmacological laboratories used simple but often unnatural animal models involving various aspects of agonistic behaviors (offense, defense, flight) to detect psycho-activity of newly synthesized drugs; the aim was not to detect “anti-aggressive” drugs as therapeutics but merely a read-out for psychoactive effects. For example, neuroleptic activity in a molecule could be easily detected using isolation-induced fighting in male mice [97]. Such models are functionally simple to run and score and therefore extremely suitable for screening, but they do not reveal the mechanism of action and do not predict the specificity of the observed effect and cannot distinguish the compound tested from any other compound that shows pharmacological activity in the model or test.

Because we had no clue about a mode of action to pursue specific anti-aggressive activity in a molecule at that time, a behavioral cascade of animal models of aggressive behavior was created. The “isolation-induced aggression” test in male mice was the primary screening test to determine an ED₅₀ (in mg/kg orally) for aggression reduction. As this measure did not reveal the specificity of the anti-aggressive activity, further tests were developed to measure the behavioral specificity of the aggression reducing effects of psychoactive drugs. By using ethological methods in mice and rats [94] and combined lead-finding and screening of more than 2000 new chemical structures, some phenylpiperazine analogues were found that fulfilled primary pharmacological criteria for a non-sedative anti-aggressive structure. In 1980, after a dedicated search for optimal anti-aggressive activity, DU27716 (fluprazine) was selected for further development. Fluprazine and its later successors (eltoprazine, batoprazine) showed the specific anti-aggressive profile in which offensive aggression was reduced, whereas social behavior and exploration were not affected. This profile has been depicted as SERENIC [98]. Eltoprazine was taken into clinical development up to phase 2B, but for several reasons, no phase 3 studies were initiated. Unfortunately, since then (1994), no new developments in the search for serenics have been undertaken.

Serenics were found and developed based on a purely translational basis: animal models of aggression predicting human (pathological) aggression [62]. Although a risky approach, no target-specific search was possible, as the putative underlying mechanisms of action were unknown. In the course of time, however, it became clear that serenics interacted with central serotonergic (5-HT) systems. In the 1980s, the rapid development of receptor binding techniques and the discovery of subtype receptors of various neurotransmitters played an increasing role in the unraveling of the mode of action of drugs. The most prominent feature of serenics was their affinity for serotonergic receptors. Over time [62] it became clear that serenics (eltoprazine) have high affinity for 5-HT₁ receptors, specifically for 5-HT_1A and 5-HT_1B receptors. Further research has shown that eltoprazine exerts its serenic activity because of its (partial) agonistic activity at 5-HT_1A and 5-HT_1B receptors. There is evidence [88, 99, 100] that serenic activity is mediated postsynaptically via 5-HT_1B receptors [84], although a role for 5-HT_1A receptors cannot be excluded [71].

Although serotonin has been considered for a long time as a very important neurotransmitter in the modulation of aggression and impulsivity [101], it does not work in isolation; other neurotransmitters clearly play an important role too [86, 102]. Apparently, however, serotonin is a key player in modulation of aggression mechanism and circuitry; a PUBMED search (March 22, 2021) on “aggression” coupled to “serotonin” yields 3010 hits, with “dopamine” 1627 hits, with “noradrenaline” 849 hits, with “GABA” 705 hits, whereas with other neurotransmitters yields lower hits.

The specific search for serenics, independent of the underlying molecular target, has not been pursued after the development stop of eltoprazine in 1994. Since then, fundamental research on aggression has dwindled for some time although many relevant studies in man and animal are still pursued. Verhoeven and Tuinier [103] pleaded for continued research into serenics, strongly supported by Coccaro et al. [101] who “hoped that new insights into the neurobiology of aggression will reveal novel avenues for treatment of this destructive and costly behavior.” The recent surge in applying new techniques in neurobiology has also brought exciting findings in the circuitry and genetics of aggression that might facilitate future search for new serenics.