Penile erection is an important and complex process involving multiple interplays between the peripheral and central nervous systems, and the endocrine system [ 1]. Dysfunction in one or more of these control mechanisms can result in impairment of the erectile process, leading to impotence. The process of penile erection is largely a haemodynamic event involving relaxation of the smooth muscle of the corpus cavernosa and arterioles. The resultant increase in blood flow into the trabecular spaces of the corpus cavernosa leads to erection. The scope of this review is to describe the peripheral and central endogenous neurotransmitters involved in mediating erectile responses and to illustrate how knowledge of the physiological mechanisms underlying erection can identify effective medical therapies for men with erectile disorders. An excellent and detailed account of the neuroanatomy, neurophysiology and pharmacology of erection can be found in recent reviews [ 2, 3].
The vascular system and smooth muscles of the corpus cavernosa with the surrounding striated muscle are controlled by nerves of the autonomic (sympathetic and parasympathetic) and somatic nervous systems, respectively, [ 4]. The neurotransmitters that are believed to play a key role in regulating the tone of the corpus cavernosum under physiological conditions are noradrenaline (NA), acetylcholine (ACh) and nitric oxide (NO). However, whilst their involvement is well established, there is also evidence implicating neuropeptides in the control of corporal smooth muscle either directly or via an interaction with cholinergic and adrenergic neurons (see [ 2] for review).
The relaxation of corpus cavernosal smooth muscle is a prerequisite for penile erection and is under the control of the autonomic nervous system. Electrical field stimulation (EFS) or muscarinic receptor stimulation induces rapid relaxation responses in precontracted human corpus cavernosum [ 5, 6]. In the case of EFS, these relaxation responses persist in the presence of atropine and guanethidine, indicating that neither cholinergic or adrenergic neurotransmitters are involved. Indeed, the relaxation responses are mediated by nonadrenergic, noncholinergic (NANC) neurotransmission. The NANC neurotransmitter responsible for mediating these relaxations of corpus cavernosa has been shown to be NO, a labile substance synthesized from l-arginine by the enzyme nitric oxide synthase (NOS) (Fig. 1) [ 7–9]. The synthesis of constitutive NO can be inhibited by arginine derivatives [ 10] making these agents extremely useful pharmacological tools for investigating the physiological functions of NO. Three distinct forms of NOS derived from different gene products have been identified and designated as inducible (iNOS), endothelial (eNOS) or neuronal (nNOS) [ 11]. The presence of nNOS has been detected in the peripheral autonomic nervous system, including the human penis [ 12]. Both nNOS and eNOS are believed to be important sources of NO in the corpus cavernosum as relaxation [ 6] and erectile responses [ 13] were reduced after removal or destruction of the endothelium. Supporting the pivotal role played by NO in mediating relaxation of the human corpus cavernosum to EFS, inhibition of NOS completely blocks smooth muscle relaxation [ 14–18]. Similarly, in rats, penile erection elicited by electrical stimulation of the pelvic nerves was blocked completely by the NOS inhibitor l-nitroarginine methyl ester (l-NAME) [ 19].
The assumption from these and other data is that, in the periphery, neuronal NO is the key if not pivotal mediator of erection. However, transgenic mice lacking nNOS are able to achieve erections and are fertile [ 20], suggesting that in these genetically altered animals compensatory mechanisms have developed. Indeed, further investigation of these animals revealed that nNOS-deficient mice have a greater abundance of eNOS in the endothelium of penile vasculature and sinusoidal endothelium within the corpus cavernosum [ 21] that seems to compensate for the absence of neuronally derived NO. Thus, inhibition of the compensatory eNOS with l-NAME prevented erectile responses in nNOS-deficient mice, thereby confirming the important role of NO as a mediator of penile erection even under these circumstances.
The relaxant actions of NO on the corpus cavernosum are caused by activation of soluble guanylate cyclase and subsequent production of cyclic GMP (cGMP) which acts as a second messenger, resulting in a decrease in intracellular calcium and thus smooth muscle relaxation [ 14]. The activity of cGMP is regulated by cyclic nucleotide phosphodiesterase enzymes (PDEs) present in the corpus cavernosum (Fig. 2). The major cGMP metabolizing PDE in human corpus cavernosum is PDE5 [ 22]. Therefore, by inhibiting the activity of this enzyme, the potent and selective PDE5 inhibitor sildenafil (Table 1), prevents the breakdown of cGMP, resulting in enhanced relaxation of the corpus cavernosum [ 23] and penile erection [ 24]. Thus, sildenafil enhances the normal physiological neural pathway responsible for penile erection (Fig. 2) which underlies its effectiveness as a treatment for patients with erectile dysfunction [ 22, 25].
Table 1. The dose required for 50% inhibition (IC50 ) for human PDE isozymes by sildenafil. PDE2, 3 and 5 were isolated from corpus cavernosum, PDE1 from cardiac ventricle and PDE4 from skeletal muscle. IC50 values were determined using cGMP (0.5 μmol/L) as substrate for PDE 1, 2 and 5 and cAMP (0.5 μmol/L) as substrate for PDE 3 and 4. Data shown are mean (SEM) (Data from )
Noradrenaline contracts the corpus cavernosum by acting on α1-adrenoceptors, as the effect is antagonized by prazosin, although there is also evidence for the presence of α2 receptors [ 5]. Similarly, prazosin effectively blocks the contractile response evoked by electrical stimulation of the nerves in the corpus cavernosum [ 5]. The role of endogenous noradrenaline released from sympathetic nerve terminals appears to be the maintenance of cavernosal and arterial tone, such that blood flow to the corpora cavernosa is maintained at a low level so that tumescence does not occur. Thus, blockade of the sympathetic drive would be pro-erectile and probably explains the tumescence-producing effect of intracavernosal α1-antagonists [ 26].
The role of endogenously released ACh in controlling erectile function in the periphery is less clear, as ACh can act in several different and indirect ways to influence smooth muscle relaxation [ 2, 3]. However, that ACh relaxes corpus cavernosum when the endothelium is present [ 5, 27], and as these effects are blocked by NOS inhibition [ 6, 28], suggests that an interaction with the NO/cGMP system is involved.
Despite the evidence implicating ACh as a neurotransmitter mediating corpus cavernosum relaxation, there is still controversy as to its physiological relevance. Experiments to investigate the potential role of ACh in erectile responses in animals have provided some evidence in favour (reviewed in [ 2, 3]), but data obtained in humans has been less convincing. Indeed, intravenous [ 29] or intracavernosal [ 26] administration of the muscarinic receptor antagonist atropine had no effect on the erectile process, suggesting that under normal conditions, endogenous ACh plays no significant role in the physiological mechanisms underlying erection. However, the role (if any) played by ACh under disease conditions where erectile dysfunction is evident remains to be established.
Many neuropeptides have been identified and shown to exert potent actions on penile tissue (reviewed in [ 3]) but the physiological role and involvement in the erectile process is unclear. The following is a review of those neuropeptides where most is known about their actions on corpus cavernosum and/or penile vasculature.
Vasoactive-intestinal polypeptide (VIP).
VIP is localized in neuronal fibres innervating the corpus cavernosum and in penile vasculature [ 3], and causes a marked relaxation of smooth muscle, including human corpus cavernosum [ 30–33]. In addition, VIP has been implicated along with NO as a NANC transmitter mediating relaxation responses in the rabbit corpus cavernosum [ 34]. However, evidence against the involvement of VIP as a NANC transmitter in human corpus cavernosum [ 35] suggests this peptide is not a key endogenous modulator of erectile function. Furthermore, whilst the relaxant activity of exogenously administered VIP was blocked by VIP antiserum, it had no effect on the relaxation induced by electrical stimulation of the nerves [ 33], suggesting that VIP is unlikely to be the NANC transmitter under these conditions. Further evidence that VIP is not the key NANC transmitter mediating erection is that intracavernosal administration of the peptide did not produce full erections in normal [ 36] or impotent men [ 37, 38]. However, use of higher doses of VIP in combination with phentolamine have been reported to induce erections sufficient for intercourse [ 39]. Thus, a role for VIP in mediating erections cannot be excluded, although its involvement as an endogenous NANC transmitter in the process remains unproved.
Neuropeptide Y (NPY).
NPY often coexists with noradrenaline in sympathetic adrenergic neurons and has been shown to act together with this amine to constrict blood vessels [ 40]. The implication is that NPY may be involved with noradrenaline in detumescence mechanisms. Although NPY is found in the erectile and penile vascular tissue of man [ 41, 42], the contractile effects on the corpus cavernosum were either variable [ 42] or absent [ 32]. In addition, NPY produced no enhancement of the contractile effect of noradrenaline on rabbit corpus cavernosum. In the absence of conclusive evidence implicating NPY in the erectile process, a neurotransmitter or neuromodulatory role for NPY remains to be shown.
Endothelin-1(ET-1) is an extremely potent contractile agent of human corpus cavernosum with ET-2 and -3 being less potent [ 43, 44]. In contrast to NPY, ET-1 enhances the contractile response to noradrenaline and may be involved in detumescence. ET-1 is localized intensely in the endothelium and to a lesser extent in the smooth muscle [ 44]. A role for ET as a neurotransmitter regulating penile erection has been proposed, involving modulation of NO.
The CNS involvement in erectile function is unquestioned but the complexity of the process means that little is known about the neuronal circuits and neurotransmitters that function under physiological conditions. However, significant roles for 5-hydroxytryptamine (5-HT), dopamine, NO and several neuropeptides have been suggested. This review will present the evidence implicating these neurotransmitters and neuropeptides in the CNS control of erectile function. The anatomical basis for these actions will not be discussed but has been reviewed recently [ 3].
The involvement of 5-HT in sexual function is complex, involving effects on erection, ejaculation and behaviour. Investigations into the physiological role of 5-HT have been complicated as many 5-HT receptor subtypes have been implicated in sexual function and 5-HT may act to facilitate or inhibit sexual responses via actions in several brain areas. Furthermore, much of the work has highlighted the existence of species differences in responses to 5-HT [ 2].
A role for 5-HT in mediating erectile responses in rats [ 45] and nonhuman primates [ 46] arose from studies using compounds that either released 5-HT (fenfluramine), or were 5-HT agonists, e.g. m-(chlorphenyl)piperazine (mcPP). Indeed, the pro-erectile effects [ 47] and reports of priapism [ 48] observed with trazodone have been attributed to mcPP produced metabolically. The identity of the 5-HT receptor responsible for eliciting penile erections in rats appears to be the 5HT2C subtype [ 49, 50] (note the change in nomenclature [ 51], as the potency of several compounds for this receptor correlated well with their ability to induce penile erections. Recent evidence supporting this proposal is that the pro-erectile effect of a selective 5-HT2C agonist was blocked by 5-HT2C but not 5-HT2B antagonists [ 52]. Whilst exogenously administered compounds may elicit penile erection, the physiological significance of these pathways remains unknown. However, the availability of selective agonists and antagonists of the 5-HT2C receptor can now be used to assess the physiological role of endogenous 5-HT in mediating erectile responses.
Dopamine has been implicated in the control of penile erection in rats [ 53] and humans [ 54, 55] by a central action mediated by D2 receptors [ 3, 56, 57]. The central locus of action of apomorphine in rodents appears to be the paraventricular nucleus (PVN) although other nearby hypothalamic structures have been implicated [ 3]. The erectile response is likely to be secondary to activation of central oxytocinergic systems and may involve central NO [ 57, 58].
In agreement with the animal data, dopamine agonists are pro-erectile in men with erectile dysfunction. Thus, apomorphine has been shown to produce penile erections in a selected group of impotent patients with no documentable organic basis for erectile dysfunction. This suggests that activation of D2 receptors may represent a means of treating psychogenic impotence [ 59]; however, the use of apomorphine is limited by side-effects such as nausea which have seriously limited its clinical use. Whilst these and other data indicate that activation of central dopamine pathways produce erections, the physiological significance of endogenous dopamine in the process remains to be shown conclusively.
The role of peripheral NO in mediating erectile responses is well established. However, in addition to this, evidence exists to suggest that central NO may be also be involved as NOS inhibition (intracavernosal) prevents 5-HT2C agonist [ 60], oxytocin and apomorphine-induced penile erections in rats [ 58]. Further evidence that central NO may be involved is that NO donors elicit penile erection when injected directly into the PVN [ 61, 62]. These reports make it tempting to speculate that a central NO/cGMP system exists in addition to that found in the corpus cavernosum and that together they are responsible for eliciting penile erection. However, the involvement of cGMP in the central action of NO is much less clear than the pivotal role that the NO/cGMP system plays in the corpus cavernosum. Indeed, injections of 8-bromo-cGMP (a stable analogue of cGMP) into the PVN, where NO elicits an erectile response, does not produce erections [ 63] suggesting that these systems may not function together in the PVN, but perhaps in different brain areas [ 3]. Further investigations of the molecular mechanisms involved in these actions of central NO will help understand if and how central NO pathways contribute to the erectile process under physiological conditions.
The central noradrenergic neurons in the brain may influence sexual function, as suggested by animal studies where yohimbine increased sexual activity in rats [ 64]. Yohimbine, an α2-adrenoceptor antagonist, has been considered to be an aphrodisiac in humans. Several placebo-controlled clinical trials have been conducted in patients with erectile dysfunction, but efficacy has been modest at best [ 65, 66]. More recently, a highly selective α2-antagonist (RS15385) has been tested for efficacy in trials for erectile dysfunction, with similarly disappointing results [ 67, 68]. These data suggest that, whilst yohimbine may have some effect on arousal in response to sexual stimuli in a subset of patients with psychogenic impotence, it is unlikely to improve erectile responses in those with broader aetiologies. Thus, α2-adrenoceptors are unlikely to play a key role in the physiological CNS pathways controlling erection.
Several neuropeptides, including opioids, oxytocin, ACTH and α-melanocyte-stimulating hormone (MSH) have been reported to influence penile erection and ejaculation in rodents. A detailed account of the complex CNS interactions that exist amongst neuropeptides is discussed by Argiolas and Melis [ 3]. Consistent with other CNS mechanisms, little is known about how these substances interact under physiological circumstances to modulate or control erectile function. However, the neuropeptides listed above clearly have central effects on penile function and the key effects are reviewed below.
The widespread distribution of oxytocin throughout the CNS suggests that, in addition to its neurohypophyseal function in parturition and milk letdown [ 69], this peptide also functions as a neurotransmitter or neuromodulator [ 70, 71]. Oxytocin has been implicated in sexual behaviour [ 72] and is a potent stimulator of penile erection after intracerebroventricular administration to laboratory animals [ 73]. The effects on erection are blocked by oxytocin-receptor antagonists [ 74] suggesting an interaction with an oxytocin receptor in the hypothalamus [ 73]. There are clearly significant interactions between oxytocin-, dopamine-and NO-driven CNS pathways involved in penile erection (see earlier). However, the physiological role that these transmitters play in the erectile process has yet to be determined, but could be evaluated using pharmacological blockade experiments.
Penile erection and related behavioural responses may be produced after central injection with ACTH and α-MSH. The effects on penile erection are highly dependent on the testosterone environment [ 75]. However, despite these and other similar observations, very little is known about their mechanism of action or physiological significance.
In contrast to many of the other neurotransmitters and neuropeptides implicated in erectile function, the opioid family of peptides are generally considered to be inhibitory to the erectile process. Evidence supporting this arises from studies where the opioid antagonist naloxone was used to investigate the role of endogenous opioids. In cats, naloxone induced erectile responses in about half the animals, suggesting that opioids may participate in an inhibitory mechanism [ 76]. In addition, opiate addicts often report impotence as a side-effect of addiction, a process reversed with naloxone [ 77]. Despite these reports, further evidence is required to confirm that endogenous opioids participate in a tonic inhibitory mechanism and that such an event is an important physiological regulator of penile erection.
In the periphery, NO appears to play a key, if not pivotal role in the erectile process. This is supported by many studies using animal and human corpus cavernosal tissue in vitro, and also animal studies in vivo. A combined feature of these studies is that blockade of NO synthesis completely prevents corpus cavernosal relaxation and penile erection. Preventing the breakdown of cGMP (the second messenger activated downstream of NO) with a specific PDE5 inhibitor such as sildenafil augments the relaxation response of the corpus cavernosa both in vitro and in vivo. Thus, inhibiting PDE5 with subsequent elevation in cGMP augments the normal physiological response to sexual stimulation and underlies the clinical efficacy observed with sildenafil in patients with erectile dysfunction. The CNS control of erectile function is more complex and less understood. While numerous neurotransmitters and neuromodulators have been shown to influence penile erection, the physiological significance of these actions remains to be answered.
Understanding the importance of NO in mediating penile erection in animals and humans is the key to identifying new medical therapies for erectile dysfunction. As our knowledge in the area of the CNS control of penile erection increases, it is likely that the importance of other neurotransmitters will be fully realized.