Advances in the study of the peripheral nervous system for erection in animals and humans

Abstract

Introduction

Since Walsh first emphasized the importance of preserving the neurovascular bundle n to protect the cavernous nerve during pelvic surgery, patients’ sexual life quality has dramatically improved. Today, nerve‐sparing radical prostatectomy is the established gold standard for organ‐confined prostate cancer patients. Recent technical advances in functional assessment such as intraoperative electrical stimulation have unveiled new anatomical features and physiological roles. Basic research has advanced understanding of cavernous nerve function, while molecular biology has uncovered the crucial role of neuronal nitric oxide in mediating erection, and has led to new treatments such as phosphodiesterase type‐5 inhibitors. A recent focus in cavernous nerve research concerns the nerve distribution external to the neurovascular bundle. The cavernous nerves in humans appear to be distributed more widely beneath the lateral pelvic fascia than in other animals, and electrical stimulation studies suggest possible involvement of these nerves in erection. These findings have prompted new surgical techniques such as the “veil of Aphrodite”, or “intrafascial nerve‐sparing” procedures.

Materials and Methods

These recent anatomical and physiological studies in humans and animals and their impact are reviewed in this article.

Conclusions

Further investigation should stimulate future advances in strategies to preserve erectile function in RP patients.

Animals

History of animal studies of erectile function

Animal research on penile erection dates from at least the latter half of the nineteenth century when Eckhard reported pelvic nerve involvement in erection in dogs [1]. Langley [2] introduced the term ‘autonomic nervous system’ and also coined the designation ‘parasympathetic’ in accord with his insights into the selective stimulatory action on fibers contained in the cranial and sacral outflows. In the twentieth century, pelvic neuroanatomic investigators had reached a consensus regarding the primary importance of the pelvic nerve plexus [3]. In 1968, technical advances enabled Lewis [4] to measure the intracavernous pressure (ICP) in bulls. Subsequently, several animal models were developed to investigate erectile function including studies in monkeys, dogs and rabbits. In the 1980s Sjöstrand [5] used plethysmography to quantify penile erection in rabbits, and Lue and co‐workers and Goldstein and co‐workers conducted experiments in monkeys, dogs and rabbits to show the hemodynamic events in the penis and the crucial role of cavernosal smooth muscle [6, 7, 8]. Epoch‐making experiments using commercially available small animals, rats and mice, were conducted in the last decade.

Walsh, the pioneer of nerve‐sparing radical prostatectomy (RP) in humans, and his co‐workers pioneered studies on penile erection using rats and their innovations enabled others to conduct erectile function studies more easily, resulting in significant progress in this research field [9]. More recently Burnett's and co‐workers reported the first experiments to use mice [10]. The addition of transgenic mice enabled experiments intended to expand our physiological and pharmacological understanding of erection [11, 12]. The neuroanatomy and physiology of the penis in rats and mice were gradually elucidated, and contemporary knowledge and technical advances will ensure further progress in this field [13].

Evaluation of erectile response in animals

Penile erection is a hemodynamic event, the consequence of relaxation of its smooth muscle, which involves both the central nervous system and local factors. The current basic concepts of erectile physiology result from in vivo and in vitro experiments, mostly performed in animals [14, 15]. In the flaccid state, the cavernous tissues are contracted by dominant sympathetic control of the arterioles and the cavernous smooth muscles [14]. Following sexual stimulation, increased parasympathetic nerve activity induces relaxation of cavernous tissue smooth muscle reducing resistance to blood flow resulting in a large influx of blood and a rapid increase in ICP. In this phase, reflex contraction of the ischiocavernous and bulbocavernous muscles also enhances the peak pressure, which rises to well above the arterial blood pressure and provides maximum rigidity to the penis. Simultaneously, the blood outflow from the cavernosum is decreased due to passive compression of the veins secondary to sinusoid expansion (venous occlusion) [16]. ICP recording is the gold standard for quantitating penile erection in animals [17]. The most common technique to induce erection is electrical field stimulation of specific sites in the nervous system responsible for erection.

Sites for electrical stimulation in animal experiments

Every site in the neural pathway responsible for erection can be used for electrical stimulation including the medial preoptic area (MPOA) in the hypothalamus, the pelvic ganglia, and the pelvic and cavernous nerves (CNs) [18]. Electrical stimulation of the MPOA was first proposed by Giuliano [19] and is thought to induce the most physiological erection because the MPOA is the central control of penile erection [18, 20].

Stimulation of the CNs to induce erection was first proposed by Quinlan [9]. The CN is the terminal part of the neural pathway of penile erection, and is located immediately proximal to the target organ, i.e., the corpus cavernosum. CNs in rats originate from the bilateral major pelvic ganglion (MPG) and mainly from the posterolateral surface of the prostate toward the corpus cavernosum (Fig. 1). In the RP nerve‐injury model, electrical stimulation of the CN itself following CN injury is technically difficult because it is necessary to stimulate proximal to the injury site. However, the pelvic ganglia and pelvic nerve located superior to the pelvic ganglia are alternative stimulation sites [21].

Figure 1.

Representative anatomical view around the MPG of rats. MPG major pelvic ganglia, PN pelvic nerve, CN cavernous nerve

ICP monitoring in rats

ICP measurement provides the most direct assessment of erectile function and makes objective evaluation of even subtle erectile responses possible, while it also facilitates a physiological approach and has expanded the understanding of penile erection. ICP is most commonly studied in rats [17]. One of the advantages of using the rat is the simple anatomy of the peripheral nervous system. The CN and the MPG, where the CN originates, are easily identified and the MPG represents the center of the pelvic nervous system (Fig. 1). The unilateral MPG in rats consists of ‘one ganglion’; it is neither a ‘plexus’ nor a ‘complex of several ganglia’, common features in other mammals [22]. Moreover, behavioral studies and other neurophysiological studies can be combined in the same study group [23]. Models of pathophysiological erectile dysfunction (ED) are easily created in rats including models of diabetes, aging, castration, CN injury, etc. [24, 25, 26, 27].

Similarities in the anatomy and neurophysiology of penile erection between rats and humans are limited compared with the reproductive organ responsible for ejaculation. Recently, the CN distribution in humans was reported to be broadly distributed around the prostate rather than confined to the neurovascular bundle [28]. In rats the ancillary nerve is as important as the main CN [29]. Sato et al. investigated the effects of unilateral and bilateral transection of the CN (main penile nerve) on the increase in ICP following MPOA stimulation in male rats (Fig. 2). Unilateral main penile nerve transection induced a 28% decrease of ICP while bilateral transection led to a 55% decrease of ICP. In other words, even after bilateral transection of the CN, significant increases in the ICP response following central stimulation were observed. The result suggested that the ancillary penile nerves, which originate from the major pelvic ganglia, have a complementary role to the CNs in the autonomic innervation of the penis.

Figure 2.

Schematic diagram of the relative contributions of the various neural pathways to the MPOA‐stimulated ICP response. Shown in parentheses are the approximate percent contribution of each penile nerve branch to the MPOA‐stimulated increase in ICP, where the ICP response elicited by the MPOA stimulation in the nerve intact condition is depicted as 100%. (adopted from the literature with permission [29])

CN injury model in animal studies

Clinically, nerve‐sparing techniques in RP dramatically improve recovery of erectile function; however, recovery is still insufficient to achieve full function [30], hence preclinical research in this field remains essential [27]. Various types of CN injury have been tested in rodents, especially in rats. These models were stratified (from less severe to more severe) by the extent of the injury, which included crushing, freezing, transecting, and excising. Table 1 shows the CN injury models in which functional assessment by ICP was performed. The least severe model is considered to be the crush model. Interestingly, however, exposure alone of the CN affected erectile function as well as the crush models [31, 32]. Neither the crush model for the nerve‐sparing RP nor the freezing model for cryoablation disrupted the nerve continuity and both preserved the nerve sheath. As an example of more severe CN injury, transection/excision techniques were used as non‐nerve sparing or nerve destruction models with a subsequent nerve graft technique.

Table 1.

Cavernous nerve injury models analyzed by ICP

Author	Type of injury	Side	Time frame
Mullerad et al. [31]	Exposure, crush, transection	Bilateral	3, 10, 28 days
Yamashita et al. [32]	Exposure	Bilateral	1, 2, 4, 8 weeks
Sezen et al. [78]	Crush	Unilateral, bilateral	1, 3, 7 days
Lagoda et al. [79]	Crush	Unilateral, bilateral	14 days
Mulhall et al. [80]	Crush	Bilateral	28 days
Muller et al. [81]	Crush	Bilateral	10 days
Cangven et al. [82]	Crush	Bilateral	7 days
El‐sakka et al. [43]	Freezing	Unilateral	1, 3 months
Paick and Lee [83]	Transection	Bilateral	0 days
Giuliano et al. [18]	Transection	Unilateral CN and lumber sympathetic chain	0 days
Sato et al. [29]	Transection	Unilateral, bilateral, main CN and ancillary penile nerve	0 days
Quinlan et al. [9]	Transection, excision	Bilateral	0, 7, 30 days
Ball et al. [84]	Excision	Bilateral	2, 4 months
Jung et al. [85]	Excision	Unilateral	1, 3 months
Burnett and Becker [86]	Transection, excision	Bilateral	28 days
Allaf et al. [87]	Transection, excision	Unilateral	14 days
Syme et al. [88]	Excision	Bilateral	3 months
Connolly et al. [89]	Excision	Bilateral	1, 3 months
Hisasue et al. [20]	Transection, excision	Bilateral	1, 3 months

CN cavernous nerve

The timing of the assessment following CN injury is important to standardize CN injury experiments. The time frames differed among the experiments (Table 1). Some of the transection/excision rats were used to investigate the immediate effect of CN injury, but most were used for the CN reconstruction with a graft or a conduit. These experiments needed a relatively longer period (around 3 months) [20, 33].

For less damaging techniques which were used as a nerve‐sparing model, a relatively shorter time frame is required. In the crush model, even animals without proerectile treatment could recover from ED over a long‐term period, generally over 2 months, thus postoperative assessment within 1 month is generally accepted (data not shown). In adulthood, each rat month is roughly equivalent to 2.5 human years [34]. Most publications indicate that erectile function recovery plateaus at 12–24 months in humans following nerve‐sparing RP [35, 36, 37], although some recent reports describe further recovery beyond 24 months [38, 39]. Therefore, 1 month (2.5 years in humans) is a reasonable time frame to assess the nerve‐sparing technique in rats [40].

Regeneration or collateral sprouting of CN and neuronal nitric oxide synthase (nNOS) in rats

Transection of myelinated fibers (sensory or motor‐nerve) results in distal fragmentation of the axon and myelin sheaths [41]. Regenerating axonal sprouts grow in association with Schwann cells attracted by a gradient of neurotrophic factors. Axonal reconnections with the periphery, formation of synaptic buttons, maturation of the nerve fiber and its reprojection to the target organ are induced. Immunohistochemistry is generally used for histological confirmation of regenerated CN fibers. Nitric oxide (NO) produced by nNOS has been identified as one of the neurotransmitters involved in the non‐adrenergic non‐cholinergic pathway related to penile erection [42]. CN injury causes a decrease in nNOS‐positive nerve fibers in the penis, and apoptosis of cavernous smooth muscle subsequently occurs [43, 44]. The penis contains sympathetic, parasympathetic, and sensory nerves; however, the presence of nerve fibers themselves does not guarantee regeneration of the erectile neural pathway, only the identification and quantification of nNOS‐containing nerve fibers confirms CN regeneration.

The functional recovery of CN: aging and neurotrophic factor

Clinically, aging prolongs functional recovery from ED following RP [30]. Seventy‐six percent of patients <60 years old experienced functional recovery, compared with only 47% of those over 65 years [45]. The mechanism, which reduces neural plasticity with aging, remains unknown. Our previous experiments in models of unilateral CN injury suggested that post‐ganglionic projecting neurons originating in the intact side would be most likely to contribute to erectile functional recovery via a plastic change such as nerve sprouting [33]. Possibly nerve regeneration is slower in aged animals and more likely to undergo secondary deterioration [46]. Axonal regeneration and collateral sprouting may be promoted by neurotrophic factors derived from the target organ [47], but possibly aging might compromise expression of neurotrophic factors and receptors [48].

Neurotrophic factors produced by the target organs are delivered to the neuronal cell bodies by retrograde axonal transport [49, 50]. Neurturin and glial cell line‐derived neurotrophic factor (GDNF) have been proposed as neurotrophic factors of penis‐projecting neurons based on the finding [51] that 98.4% of penis‐projecting neurons in the pelvic ganglia were positive for GDNF family receptor alpha‐2 (GFRa2), a neurturin receptor. Moreover, GFRa2 is closely related to the development of nNOS‐positive neurons, and nNOS‐positive nerve fibers are diminished in GFRa2 knockout mice. Therefore, it is speculated that neurturin retrogradely supports nNOS‐positive neurons via GFRa2 [52]. Change with age in these conditions might adversely alter the neural plasticity needed for recovery from ED in RP patients.

Aging may be associated with decreases in nNOS‐positive neurons in the MPG and nerve fibers in the penis thereby impairing erectile function [53, 54]. The age‐dependent decrease of nNOS‐positive neurons found in our previous study was connected to a change in GFRa2 expression, which was induced by the interaction between nNOS and the neurturin−GFRa2 pathway. This outcome may be one of several factors underlying age‐related ED.

Humans

Discovery of an excellent landmark regarding the CN in humans

Walsh and Donker [55] first described the human anatomy basis for erection in 1982. ED following RP reflects injury to the pelvic nerve plexus which is the source of autonomic innervation to the corpus cavernosum. Recognition of the neurovascular bundle (NVB) including the CN is one of the great accomplishments of urological surgery and led to nerve‐sparing RP [30, 56], which today is an essential component of RP and dramatically improves postoperative erectile function [57]. Despite this major advance, post‐RP ED rates still vary widely from 15 to 91% [58]. Usually erectile function after RP gradually recovers after a period of complete absence of spontaneous erection. Consequently surgeons still face the challenge of improving the recovery rate and shortening the duration of ED; goals which may require new approaches to rehabilitation [59] and novel nerve‐sparing techniques [60].

New observations on the course of the CN

Initially it was believed that the CN ran adjacent to the NVB in the posterolateral aspect of the prostate from the base of the prostate to its apex. Subsequently it was determined that the CN joined the NVB at a point distal or inferior to the bladder−prostate junction (Fig. 3) [48, 61]. A recent 3D computer‐assisted anatomical dissection study confirmed that CN fibers do not strictly follow the NVB course, but are distributed at several levels, in a fan‐like formation [62]. Other recent reports also suggest that the periprostatic CN is more diffuse than initially believed [48, 49, 61]. In patients with no definite NVB formation observed on magnetic resonance imaging, the nerves associated with erectile function may run along both sides of the prostate and spread more anteriorly than occurs when the NVB is more clearly observed [63]. During early fetal development the fibers of the CN enclose the prostatic and membranous urethra dorsally and laterally. As the prostate grows, the CNs running beside the prostate become displaced further anteriorly and spread, thus forming a concave shape (like a ‘curtain’) of the NVB [64]. Branches are distributed around the prostate on the prostatic fascia outside of the NVB, and possibly contribute to normal erectile function. Based on this possibility, new nerve‐sparing techniques such as the ‘veil of Aphrodite’ to preserve this function were proposed [60]. We performed immunohistochemical staining with nNOS‐antibody and divided the prostatic hemisphere into 6 zones to assess the distribution of nNOS‐positive nerve fibers at the apex, mid‐portion, and base of the prostate. Sixty‐five percent of nNOS‐positive nerve fibers were distributed in a 3–5 O'clock sextant (P < 0.001).

Figure 3.

The scheme for the cavernous nerve course from the previous literature. In humans, cavernous nerves originate from the pelvic ganglia, and run posteriorly far from the NVB especially at the base of prostate level. Dotted lines represent the recommended incision line for ‘veil of Aphrodite’ (upper) and ‘the line for appropriate preservation of cavernous nerve’ (lower)

The definitive function of nerves outside of the NVB

Recently interest has focused on the CN exterior to the NVB with the belief that improved knowledge might lead to better preservation of pro‐erectile nerves. Conventional nerve‐sparing RP makes it possible to preserve around 55% of all periprostatic nerve fibers [65]; however, the definitive function of nerves exterior to the NVB remains controversial. Immunohistochemical studies with S‐100, a general neural marker, revealed that 30–40% of nerve fibers were exterior to the NVB [66].

Increased intracavernous or intraurethral pressures elicited by electrical stimulation of the anterior [28] and posterior surfaces [67] of the prostate indicate the existence of pro‐erectile nerve fibers outside of the NVB; however, these studies were performed in a limited number of patients and intraurethral pressure might not reflect true erectile function. In a detailed immunohistochemical study, Costello et al. [68] recently reported conflicting results in that only 11.1% of the nNOS‐positive nerve fibers were located on the anterior half of the prostate, and that most of the anterior nerves were sympathetic or somatic (Fig. 4). Furthermore, they also suggested that nNOS‐positive nerves on the anterior surface were most probably innervating the prostatic stroma. Indeed, NO synthase and phosphodiesterase type‐5 are reported to express in the prostate [69, 70]. Possibly, the anterior nerves contribute to smooth muscle relaxation in the prostate, but not in the penis. We found anterior fibers, but unlike the NVB region, they did not show the assumed compensatory increase in numbers in preoperative ED patients [71]. Further investigations will be necessary to clarify the definitive role of the nNOS‐positive nerves on the anterior aspect of the prostate.

Figure 4.

The scheme for the nerve distribution around the prostate. Left diagram shows the fiber distribution stratified by the nerve function based on the immunohistochemistry results. Right diagram shows the nNOS‐positive nerve fiber distribution around the prostate. The diagrams were made from the data adopted from the results by Costello et al. [68]

Existing clinical features and CN distribution

There is considerable individual variation in CN distribution although the reasons for this variation remain unclear [66]. There is no evidence to suggest that any preoperative clinical features influence the CN quantity. Rabbani et al. [45] suggested that better baseline erectile function and younger age convey a better prospect for erectile recovery. Others have suggested that prostate size does not influence the periprostatic fiber quantity [72].

Recently, we investigated 46 hemispheres of 23 non‐nerve‐sparing RP patient specimens [71]. We used a multivariate analysis to identify possible predictive factors for the quantity of overall nNOS‐positive nerve fibers. We included patient's age, prostate size (specimen weight), body mass index, international index of erectile function, and objective erectile function. We found no influence of these parameters on fiber quantity without the objective erectile function. The fiber count was greater in the objective ED group (1500: 382–2760) than in the non‐ED group (649: 156–2916) (P = 0.009). Thus, baseline erectile function greatly impacts CN quantity and distribution.

A possible explanation for the increased CN fibers in ED patients is collateral sprouting of nNOS‐positive nerve fibers, which might occur in the nNOS‐positive ganglionic neuronal soma located in the pelvic nerve plexus and/or on the prostate. In support of this hypothesis, nNOS‐positive nerve fibers in ED patients were significantly thinner than non‐ED patients [71]. Furthermore, the ganglionic neurons were recently suggested to distribute around the prostate [18, 66]. We found that CN quantity was similar at the base, but compared with non‐ED patients greater at the mid and apex zones in the ED patients; whereas quantity and distribution of nNOS‐positive ganglionic neurons did not differ. Chronic renal failure rats had significantly higher levels of nNOS mRNA and protein in the MPG and penile tissues than controls which might represent a compensatory increase of nNOS fibers to recover penile erectile function [73]. The impaired baseline erectile function which is related to damaged cavernous tissue or nerve terminals induced the collateral sprouting of nerve fibers around the prostate; this reaction being more marked in the distal portion [74].

The prostatic apex is the key anatomical region for NVB preservation. In our study, the fibers increased at the apex and are possibly thinner in ED patients [75]. Therefore, gentle and meticulous preservation, such as an early release of the NVB, is essential [76, 77]. This approach has been previously reported to provide better functional recovery, raising the possibility of traction injury and/or poor visualization [76]. Moreover, our study showed that more nNOS‐positive nerve fibers spread over the NVB to the 5–6 O'clock region in ED patients than in non‐ED patients [75]. Previously, in an intraoperative electrical stimulation study, Takenaka et al. [67] reported that posterior periprostatic CNs are responsible for erection, thus supporting the view that CN preservation in this area might contribute to preserving erectile function especially for patients with preoperative ED.

Conclusions

The CN is the essential autonomic peripheral nerve for normal erection, and its anatomical distribution is complex in both animals and humans. Its pathway in humans is not confined to the NVB and the CN runs posteriorly far from the NVB especially at the base of the prostate level, and forms a fan‐shaped distribution. Regarding the nerve fibers on the anterior surface, the nNOS‐positive fibers were few and had little influence on the erectile function. Furthermore, ED from chronic disease status may induce the compensatory increase of CN around the NVB. Studies on the peripheral nervous system with regard to erection in both animals and humans significantly advanced knowledge of the anatomical and physiological roles of the CN. Further investigation should stimulate future advances in strategies to preserve erectile function in RP patients.