Autonomic-somatic communications in the human pelvis: computer-assisted anatomic dissection in male and female fetuses

Abstract

Sphincter continence and sexual function require co-ordinated activity of autonomic and somatic neural pathways, which communicate at several levels in the human pelvis. However, classical dissection approaches are only of limited value for the determination and examination of thin nerve fibres belonging to autonomic supralevator and somatic infralevator pathways. In this study, we aimed to identify the location and nature of communications between these two pathways by combining specific neuronal immunohistochemical staining and three-dimensional reconstruction imaging. We studied 14 normal human fetal pelvic specimens (seven male and seven female, 15–31 weeks’ gestation) by three-dimensional computer-assisted anatomic dissection (CAAD) with neural, nitrergic and myelin sheath markers. We determined the precise location and distribution of both the supra- and infralevator neural pathways, for which we provide a three-dimensional presentation. We found that the two pathways crossed each other distally in an X-shaped area in two spatial planes. They yielded dual innervation to five targets: the anal sphincter, levator ani muscles, urethral sphincter, corpus spongiosum and perineal muscles, and corpora cavernosa. The two pathways communicated at three levels: proximal supralevator, intermediary intralevator and distal infralevator. The dorsal penis/clitoris nerve (DN) had segmental nitrergic activity. The proximal DN was nNOS-negative, whereas the distal DN was nNOS-positive. Distal communication was found to involve interaction of the autonomic nitrergic cavernous nerves with somatic nitrergic branches of the DN, with nitrergic activity carried in the distal part of the nerve. In conclusion, the pelvic structures responsible for sphincter continence and sexual function receive dual innervation from the autonomic supralevator and the somatic infralevator pathways. These two pathways displayed proximal, intermediate and distal communication. The distal communication between the CN and branches of the DN extended nitrergic activity to the distal part of the cavernous bodies in fetuses of both sexes. These structures are important for erectile function, and care should therefore be taken to conserve this communication during reconstructive surgery.

Introduction

Urination, defecation and sexual function are under the control of the somatic and autonomic nerve pathways, which involve reflex loops supported by neural loops surrounding the levator ani muscles (LAM). Classically, two distinct compartments in the lower pelvis are recognised, separated by the LAM, and innervated by two nerve pathways (Benoit et al. 1999). The supralevator pathway mostly involves autonomic nerves: the superior hypogastric plexus (SHP) with hypogastric nerves (HN), the pelvic splanchnic nerves (PSN) and the inferior hypogastric (pelvic) plexus (IHP) with its terminal projections. This plexus has been the subject of various anatomical descriptions, which have tended to adopt a systematic approach to the complex distal branches (Fritsch, 1989; Colleselli et al. 1998; Baader & Herrmann, 2003; Mauroy et al. 2007a; Alsaid et al. 2011; Moszkowicz et al. 2011a). The infralevator pathway consists of somatic nerves: the pudendal nerve (PN) and its terminal branches. There is less debate about the distribution of nerves in the somatic system than in the autonomic system, but various descriptions of the somatic innervation of the LAM have been reported (Roberts et al. 1988; Narayan et al. 1995; Wallner et al. 2008).

The dorsal penis/clitoris nerve (DN) has been shown to contain autonomic nerve fibres, suggesting that there may be communicating branches at several levels of the pelvis and perineum (Yucel et al. 2004; Martin-Alguacil et al. 2008a). Many descriptions of the pathways followed by these communicating branches have been the published, based on gross anatomical investigations and the micro-dissection of IHP terminal branches and PN branches in adult specimens (Paick et al. 1993; Colombel et al. 1999; Moszkowicz et al. 2011b). More recently, specific neuronal immunohistochemical staining and three-dimensional reconstruction imaging techniques have facilitated efforts to determine the topography and nature of these branches in fetal specimens (Yucel & Baskin, 2003).

The existence of these autonomic-somatic communicating branches suggests possible plasticity in the supply of nerves for clitoris/penis erection, bladder and urethral and anal sphincter contraction/relaxation. This plasticity is of potential clinical interest for the rehabilitation of postoperative uro-genital and digestive dysfunctions and their pharmacological treatment.

The aim of this study was to improve the anatomical description of the infra- and supralevator pathway branches and the autonomic-somatic communications of these branches, by performing an immunohistochemical analysis with three-dimensional reconstruction.

Materials and methods

This work was carried out according to a specific protocol submitted to and approved by the French Biomedicine Agency. The fetal specimens were obtained from miscarried fetuses or fetuses aborted legally, and authorisation for the scientific use of this material was obtained from the parents. The fetuses had no maceration or macroscopic abnormalities on macroscopic pathology examination.

We studied 14 fetuses, seven male and seven female, with a crown-rump length (CRL) of 110–310 mm. All fetuses studied had a gestational age between 15 and 31 weeks. The gestational age of each fetus was determined from the CRL and fetal heel-to-toe length, corrected for the first-trimester ultrasound CRL measurement, and was confirmed, on post mortem examination, by estimating organ maturation (Hern, 1984).

The entire pelvis, from the sacrum to the pubic arch, was removed en bloc. Organs were fixed by incubation in formalin (10% formaldehyde) for 48 h. We then cut the tissue into transverse slices at 4-mm intervals. The tissue slices were placed in baskets, processed and embedded in cardboard moulds filled with paraffin. We then cut series of 5-μm-thick sections at 50–150-μm intervals with a sliding microtome. In total, we obtained 150–320 sections for each fetus.

We investigated intrapelvic and perineal innervation by conventional histological [haematoxylin-eosin (HE)] and immunohistochemical staining of neural structures, as previously described (Alsaid et al. 2009). All immunohistochemical and histological staining procedures were performed manually by the authors. Briefly, the neuronal markers used were detected with polyclonal antibodies against protein S100, the neural isoform of nitric oxide synthase (nNOS) and peripheral myelin protein (PMP 22) (Table 1). The nNOS is generated in human erectile bodies and is mostly found in the cavernous tissue of the clitoris and the penis, particularly in the pro-erectile nerve bundles and vascular and sinusoidal endothelium (Burnett et al. 1992; Hoyle et al. 1996; Martin-Alguacil et al. 2008b). The nNOS antiserum used recognises a peptide of 155 kDa on Western blots of mouse brain (quality control data supplied by the manufacturer) (Baizer & Broussard, 2010). PMP 22 is a 22-kDa glycoprotein produced by myelinating Schwann cells in the compact myelin of the somatic peripheral nervous system (Snipes et al. 1992; Bremer et al. 2010). The avidin-biotin-peroxidase detection procedure was carried out with the Vectastain ABC kit (ref. PK6100; Vector Laboratories, Burlingame, CA, USA). Chromogenic detection was performed with the DAB detection kit (DAB, ref. SK-4100; Vector Laboratories). As a control for all immunohistochemical analyses, we used non-immune serum or IgG at an equivalent dilution.

Table 1.

Primary polyclonal antisera

Antigen	Dilution	Incubation time	Incubation temperature	Antigen retrieval	Positive control	Species	Provider	Code
S-100	1/400	30 min	Room	No	Human sciatic nerve	Rabbit	Dako; DK	Z0311
nNOS	1/200	12 h	4 °C	No	Human adult cavernous cnerve	Rabbit	Cayman; USA	160870
PMP22	1/100	12 h	4 °C	No	Human sciatic nerve	Rabbit	Abcam; USA	Ab61220

Serial stained and immunolabelled two-dimensional sections were used for three-dimensional reconstructions. Analyses of HE-stained sections under appropriately high magnification (×4 to ×40) made it possible to identify the various anatomical structures (organs, bones and fascia). Subsequent sections, treated with an antibody against S100, were used to identify pelvic-perineal nerves and communicating branches. By comparing HE-stained sections with sections stained with antibodies against S100, PMP22 and nNOS, we were able to determine the nature of the nerve fibres identified: somatic or autonomic and nitrergic (erectile). The sections were taken at almost the same level, with a negligible interval between sections (5 μm). The computer system comprised a personal laptop computer (Windows XP) equipped with the Epson Perfection V750 digitisation system, silverfast AI digitisation software, adobe photoshop image processing software and surfdriver software for Windows (winsurf image reconstruction software, version 4.3). All the sections were digitised by direct scanning at a resolution of 4800 dots per inch (dpi), and then the images were then stacked and aligned. The brightness and contrast of histological tissue images were adjusted in adobe photoshop. The pelvic anatomical structures and nerve fibres were outlined manually on all histological sections. A three-dimensional analysis of the location, course and distribution of the nerve fibres was then carried out along the x- and y-axes to generate a movie.

Results

Supralevator pathway

The IHP acted as a four-sided integration centre for the autonomic supralevator pathway. It received two groups of afferent branches: the HN from the SHP and the PSN from two to four sacral roots. The upper part of the IHP put out branches to the upper part of the rectum and the neck of the bladder. The lower part of the IHP gave rise to efferent branches running in three distal directions: postero-inferior, lateral and antero-inferior (Fig. 1B).

Fig. 1.

(A,B) Three-dimensional computer-assisted anatomic dissection from transverse immunolabelled histological sections of a 15-week-old male fetus. (A) Lateral view of intrapelvic organs showing the infralevator neural pathways of the pudendal nerve (PN) behind the levator ani muscles (LAM) and its terminal branch [dorsal nerve of the penis (DNP)]. (B) Same view with transparency of the LAM, showing the distal distribution of the inferior hypogastric (pelvic) plexus (IHP), in three directions: postero-inferior for the anal sphincter, lateral for the LAM, and anterior-inferior for the neurovascular bundles (NVB), which travel posterolaterally to the prostate (P), nerve fibres from the NVB forming three projections: anterior for the urethral sphincter (US), antero-lateral cavernous nerve (CN) for the corpora cavernosa (CC) and postero-lateral spongious nerves (SN) for the corpus spongiosum (CS). (C) Same lateral view with the two pathways, (D) Schematic diagram of the supra- and infralevator pathways. These two pathways ensure the dual innervation of five targets: (1) anal sphincter (AS); (2) LAM; (3) urethral sphincter (US); (4) CS and perineal muscles; and (5) CC. Communications between the two pathways occurred at three levels (blue arrow in D): proximal communication, intralevator communication (IC) and distal communication (DC). The two pathways crossed distally in an X-shaped area in two spatial planes (cycle in C,D).

The postero-inferior branches travelled toward the rectal wall and the smooth anal sphincter. Fibres in lateral directions branched out toward the LAM. Nerve fibres from the IHP converged anteriorly and distally to form the neurovascular bundles (NVB) with the adjacent vessels. The NVB descended at the 4–5 and 7–8 o'clock positions to the prostate in males or to the vagina in females. Some NVB nerve fibres branched medially to innervate the seminal complex (prostate, seminal vesicles and vas deferens) or the vaginal wall. More caudally, the NVB gave rise to nerve fibres following three major projections: an anterior projection for the urethral sphincter complex, an antero-lateral projection [cavernous nerves (CN)] travelling anterolaterally to the prostate/vagina to reach the corpora cavernosa, and a postero-lateral projection [spongious nerves (SN)] continuing posterolaterally to the prostate/vagina to innervate the corpus spongiosum.

Efferent distal branches from the lower part of the IHP tended to radiate out in five main directions: (1) internal anal sphincter; (2) LAM; (3) bladder neck and internal urethral sphincter; (4) seminal complex/vagina and corpus spongiosum; and (5) corpora cavernosa (Fig. 1C,D).

Infralevator pathway

The PN and its branches were involved in the infralevator pathways. The PN originated at the sacral plexus. It passed between the piriformis and coccygeus muscles and crossed the spine of the ischium, subsequently following the same course as the internal pudendal vessels in the pudendal canal (Alcock's canal).

The LAM received branches from the infralevator pathway. These branches left the PN at the level of spine of the ischium, before entering the pudendal canal. The PN gave off the inferior rectal nerves, two sets of which (posterior and lateral) penetrated the external anal sphincter. The external urethral sphincter also received different branches of the PN arising just before its penetration of the urogenital diaphragm. Terminal branches of the PN were grouped together in two nerves: the perineal nerve to the perineal muscles and the dorsal nerve of the penis/clitoris (Fig 1A).

Branches from the PN extended in five main directions: (1) external anal sphincter; (2) LAM; (3) external urethral sphincter; (4) bulbospongiosus muscles and corpus spongiosum/vestibule; and (5) ischiocavernosus muscles and corpora cavernosa of the penis/clitoris (Fig. 1C,D).

The supralevator-autonomic and infralevator-somatic pathways displayed proximal, intermediate and distal connections.

Proximal communications

Some fibres branched out from the PN just after its separation from the sacral plexus and before it reached the course of the infralevator pathway, joining the supralevator branches of the IHP via the lateral face of its posterior portion and intermingling with them. These branches constituted proximal communications between the infra and supralevator pathways. The number and density of these fibres were variable, but they were consistently present, on both sides, in both male and female fetuses (Fig. 2). These somatic (PMP 22-positive) fibres were non-nitrergic (nNOS-negative) in both sexes and on both sides.

Fig. 2.

Serial transverse sections of 21-week-old female (A,B) and 20-week-old male (C,D) fetuses, at 150-μm intervals, immuno-stained with antibody against S100 and scanned at an optical resolution of 4800 dpi, showing the microscopic (×4) appearance of the nerve fibres (frame in A,B). (E) Three-dimensional computer-assisted anatomic dissection from transverse immuno-labelled histological sections of the male fetus showing the two levels of section in (C,D). Nerve fibres leave the pudendal nerve (PN) (A,C arrowhead) to reach the inferior hypogastric (pelvic) plexus (IHP) via the lateral face of its posterior portion (B,D arrowhead) (SN, sciatic nerve).

Intermediate communications

Lateral fibres from the IHP travelled laterally to the parietal pelvic fascia, entering the LAM and continuing distally between the iliococcygeus and the pubococcygeus muscles. Most of these fibres dispersed in the LAM, but we were able to follow some, with great care, to more distal positions. They split off from the inferior border of the LAM and continued in the direction of the pudendal pedicles (Fig. 3). This situation was observed in two male subjects, with a 50-μm interval between sections used. The observed fibres were autonomic (PMP 22-negative) and non-nitrergic (nNOS-negative).

Fig. 3.

(A,B) Microscopic appearance (×4) of serial transverse sections of 15-week-old male fetus (50-μm intervals) at the level of membranous urethra, immuno-stained with antibody against S100. (C) Three-dimensional computer-assisted anatomic dissection showing the two levels of section in (A,B). Nerve fibres from the inferior hypogastric (pelvic) plexus (IHP) travelled distally in the levator ani muscles; some fibres (arrowhead in A) cross the lower border of the muscle in the direction of the pudendal pedicle (Pud. Ped.) (arrowhead in B).

Distal communications

The distal communications occurred at the level of the penile/clitoral root. The supralevator CN travelled along the antero-lateral surface of the urethral sphincter to reach the medial surface of the crural bodies. They then dispersed in the penile/clitoral hilum accompanying the cavernous arteries. The CN were strongly nitrergic (nNOS+) and their autonomic non-myelinated nature was confirmed by a lack of staining for myelin (PMP22−) (Fig. 4).

Fig. 4.

(A,E) Transverse sections at the level of ischiopubic root (IPR), showing the clitoris and the penile hilum in 26-week-old female and 20-week-old-male fetuses, respectively. Sections immunostained with antibody against S100 and scanned at an optical resolution of 4800 dpi. Microscopic appearance of the dorsal nerve of the clitoris (DNC) in B (from inset in A) and dorsal nerve of the penis (DNP) in (F) (from inset in E) and their surrounding nerve fibres, treated with an antibody against nitrergic fibres (anti-nNOS) (in C and G) and an antibody against the myelin sheath (anti-PMP22) (in D and H). At the proximal part of the corpora cavernosa (CC) in both sexes, nerve fibres branching from the DN (back arrow in B, F) had strong nitrergic activity (white arrowhead in C, G) in both sexes. The proximal portion of the DN had no nitrergic activity at this level (white star in C, G). These branches were somatic (PMP22+, black arrow in D,H) and had the same somatic nature as DN (PMP22+, black star in D,H). The cavernous nerves (CN) directed to the corpora cavernosa (CC) were strongly nitrergic (black bold arrow in C,G) and autonomic and non-myelinated in nature (PMP22−, white bold arrow in D,H).

The infralevator PN became the dorsal nerve (DN) of the penis or clitoris after it had crossed the urogenital diaphragm and reached the crural bodies. The somatic DN (PMP22+) had segmental nNOS activity (Fig. 5). Its proximal portion, branching off from the PN, was nNOS-negative. However, a weak nNOS signal was obtained for the most distal portion, beyond the top. Dorsal nerve branches on the dorsal side of the corpora cavernosa distal to the pubic arch were positive for the marker of nitrergic activity.

Fig. 5.

(A,D) Schematic diagrams of pelvic transverse sections at the ischio-pubic root (IPR) in female and male fetuses, (B,C) Serial transverse sections of 26-week-old female fetuses at the level of the clitoris hilum (frame in A) immuno-stained with antibody against S100 in (B) and antibody against nNOS in (C) and scanned at an optical resolution of 4800 dpi. (E–H) Microscopic appearance (×10) of the dorsal nerve of the penis (DNP) in 20-week-old-male fetuses; proximal portion near the penile hilum (black frame in D) and distal portion in the penis (grey frame in D), immuno-stained with antibody against S100 in (E,G) and antibody against nNOS in (F,H). The cavernous nerves (CN) travel antero-lateral to the urethra (U) and reach the corpora cavernosa (CC). CN gave a positive signal with anti-nNOS antibodies in both sexes (black arrow in C,F), The somatic dorsal nerve of the clitoris (DNC) and the DNP had segmental nitrergic activity and their proximal portions were negative for nNOS (white arrowhead in C,F), with parietal anti-nNOS staining observed in the distal portions (black arrowhead in C,H).

Two to three nerve fibres branching from the DN that were found to be PMP22+, confirming their somatic nature, displayed strong nitrergic activity (nNOS +) in both sexes (Fig. 4). The change in the nitrergic nature of the DN related to these retrograde branches: the DN became nNOS-positive when the CN interacted with DN branches (Fig. 6).

Fig. 6.

(A) Computer-assisted anatomical dissection (CAAD) of the pelvic structure with pubis transparency in a 26-week-old female fetus, shown in antero-lateral view. (B) Zoomed view (frame in A) without the bone. (C) Same view with nerve fibre transparency, with nitrergic nerve fibres shown in pink. The distal branches from the inferior hypogastric (pelvic) plexus (IHP) formed the neurovascular bundles (NVB). Some nerve fibres converged in an anterior position to innervate the urethral sphincter (US). The NVB continued its postero-lateral course to the vagina, to reach the corpus spongiosum vestibular bulb via the spongious nerves (SN). The cavernous nerves (CN) travelled anterolaterally to the vagina (V) and reached the corpora cavernosa of the clitoris (C). All the CN fibres were nitrergic. The pudendal nerve (PN) became the dorsal nerve of the clitoris (DNC), which had segmental nitrergic activity, with a non-nitrergic proximal portion (white arrowhead in C) and a nitrergic distal portion (black arrowhead in C). The change in the nitrergic nature of the DN related to the DNC branches (black arrow in B,C), which interacted with the nitrergic CN and carried the nitrergic activity to the DN. This interaction occurred in the X-shaped region of communication between the infra- and supralevator pathways.

With our three-dimensional reconstruction, it was possible to represent the supra- and infralevator pathways on both sides of the LAM and their relations. The infralevator PN pierced the urogenital perineal diaphragm and ascended to reach the cavernous bodies, whereas the supralevator CN, when dispersing in the penile/clitoral hilum, tended to follow a descending course. An X-shaped area, in two spatial planes, in which the two pathways crossed was noted (Fig. 1C,D). Distal communication between the somatic DN branches and the autonomic CN occurred at this cross-over point, with the DN carrying nitrergic activity distally (Fig. 6).

Discussion

This study provides information about the dual innervation of pelvic structures by the autonomic supralevator and somatic infralevator pathways and the communications between these pathways, based on the computer-assisted anatomic dissection (CAAD) of both male and female human fetuses.

At the end of the 8 weeks of gestation (Carnegie stage 23), the neuro-anatomical topography of the fetus resembles that in adults (Fritsch, 1989; Arango-Toro & Domenech-Mateu, 1993; Alsaid et al. 2011). Moreover, the immunohistochemical characteristics of the penile dorsal nerve in fetal specimens were consistent with the results obtained for adults (Yucel & Baskin, 2003). The fetus is considered an anatomically and physiologically good experimental model for studies of pelvic-perineal innervation.

We have previously described the concept of the supra- and infralevator pathways (Benoit et al. 1999), with the IHP and its efferences ensuring the autonomic innervation of the anatomical structures above the LAM. The PN and its branches are located below the pelvic diaphragm and supply the somatic impulse toward the perineum and the external genital structures. Interaction between the two systems ensures a normal micturition/defecation-continence cycle and normal sexual function. The efferences from the IHP have been the subject of various descriptions based on anatomic dissection (Baader & Herrmann, 2003; Costello et al. 2004; Mauroy et al. 2007b). However, we subdivided the distal branches of the lower part of the IHP into three directions: postero-inferior for the rectum and anal sphincter, lateral for the LAM, and antero-inferior for the prostate/vagina, the urethral sphincter and the erectile bodies.

The PN gives off the inferior rectal nerve, which extends to the external anal sphincter muscle, the perineal nerve, which extends to the perineal muscles (bulbospongiosus muscles and ischiocavenosus muscle), and the dorsal nerve of the clitoris/penis. Somatic innervation of the external urethral sphincter has been reported before, with some authors claiming that this innervation is generated by a supralevator somatic branch (Karam et al. 2005) that may originate from the sacral plexus. Others have reported this somatic innervation results from the PN crossing the urogenital diaphragm or from perineal branches of the PN (Junemann et al. 1987; Narayan et al. 1995). The somatic system is easier to describe than the autonomic system, but several studies on cadavers have yielded conflicting results concerning the innervation of the LAM by branches of the PN (Roberts et al. 1988; Shafik et al. 1995; Barber et al. 2002) or by the levator ani nerve (LAN) (Wallner et al. 2008), which arises directly from the sacral plexus.

Dual autonomic supralevator and somatic infralevator innervation occurred in five directions, playing a role in urination, defecation and sexual function. Innervation of the anal sphincter, the urethral sphincters and the LAM is responsible for micturition-continence and defecation cycles. Innervation of the other two structures is of importance in ejaculation/orgasm (seminal complex, vagina, corpus spongiosum and bulbospongiosus muscles) and in erection (corpora cavernosa and ischiocavenosus muscles) functions. This description of this dual innervation improves our understanding of the reflex loops around the LAM and demonstrates the importance of autonomic nerve fibre preservation during surgical procedures, as these fibres play an important role in pelvic functional cycles.

Autonomic-somatic communications were found to occur between the supralevator-autonomic and the infralevator-somatic pathways at several levels. Proximal communication between the supra- and infralevator pathways has been reported before (Arango-Toro & Domenech-Mateu, 1993; Benoit et al. 1999) and our description of some fibres of the PN joining the supralevator course of the IHP and intermingling with it are consistent with these previous studies.

There have been reports of intermediate intralevator communication. Wallner et al. (2008) reported a somatic communicating nerve branch between the LAN and the PN in more than half the hemi-pelvises studied. In our study, the LAN was not identified distinctly. The only branches identified in relation to the sacral plexus were the proximal infra-supralevator communication between the PN and the IHP. No distinct pathway was observed when these communicating branches, which intermingle with the autonomic plexus, were followed. We observed some nerve fibres branching off from the IHP toward the LAM and travelling laterally to the pelvic fascia. Most were dispersed in the lower part of the muscle, but some followed an intralevator course to the emergence of the pudendal pedicle. These fibres were very fine and were observed in only two subjects. These fibres were considered to constitute intermediate communications between the autonomic supralevator and the somatic infralevator pathways, partly accounting for the autonomic innervation of the proximal part of corpora cavernosa. Given the small number of subjects concerned, further studies are required to determine the occurrence of the intralevator communications and to clarify their nature.

Various reports have considered the distal communicating branches between the cavernous and dorsal nerves and their potential importance in erectile function. In a previous dissection study (Colombel et al. 1999), we identified communicating branches from the CN that joined the DN at an acute angle but did not determine their nature. Paick et al. (1993) also reported that the CN sent connecting branches to the DN or completely merged into the DN. They suggested that the DN acts as a ‘carrier’ of the CN to the distal corpora cavernosa and glans. More recently, Yucel & Baskin (Yucel & Baskin, 2003; Yucel et al. 2004) demonstrated segmental nitrergic activity of the DN in both sexes. They used three-dimensional reconstruction to identify branches from the cavernous nerve joining the dorsal nerve of the penis and sending out nNOS-positive fibres in males. In females, they put forward an explanation for the loss of nNOS positivity in the dorsal nerve fibres of the clitoris proximal to the clitoral body hilum, involving possible communication between nNOS-immunoreactive CN and the distal dorsal nerve of the clitoris. In our study, based on similar methods, we confirmed the occurrence of segmental nitrergic activity in both sexes. This conversion was related to the interaction between the CN and the DN in the hilum of the penis/clitoris. We also highlighted the existence of some somatic nitrergic DN branches carrying the nitrergic reaction from the CN to the distal part of the DN. These observations are consistent with the description of the DN as a carrier of autonomic nitrergic activity to the distal part of the cavernous tissues (Paick et al. 1993; Yucel & Baskin, 2003; Yucel et al. 2004). The nerves in the corpora cavernosa are known to extend from the CN. However, we observed that the distribution of the CN was limited to the penile/clitoris root and the proximal part of the corporeal bodies, with the DN supporting innervation to more distal areas. The DN is considered to be the nerve conferring sensitivity on the external genital organs, the glans skin in particular. The DN is not considered an autonomic nerve, but it contains autonomic nerve fibres supplying distal nitrergic reactivity, highlighting the importance of the DN in erection physiology.

This hypothesis has significant implications for surgery and improves our understanding of erectile dysfunction after pelvic surgery, radical prostatectomy in particular. It suggests possible plasticity in the supply of nerves for penis erection. The notion of segmental DN nitrergic activity is important during sex reassignment surgery and reconstruction of the external genital. In clitoroplasty or phalloplasty, the mobilisation of the glans with its NVB (Acimi, 2008) should respect the course of the antero-lateral cavernous nerve. In transsexual female-to-male surgery, the precise location of the recipient nerves must be determined distally to the X-shaped area of communication, to preserve nitrergic activity in the distal part of the erectile bodies.

Sacral neuromodulation (SNM) has recently shown clinically important benefit for the sexual function of both genders, mainly in individuals with lower urinary tract symptoms (Pauls et al. 2007; Lombardi et al. 2008). The mechanism by which SNM works is not completely known (Bernstein & Peters, 2005) but one potential explanation is its action on the parasympathetic system that influences both detrusor contraction and erectile function through somatic spinal loops (Leng & Chancellor, 2005). Autonomic manifestations of sexual function improvement might involve the distal caverno-pudendal nitrergic communications.

Conclusion

The pelvic structures responsible for sphincter control and sexual function receive dual innervation from the autonomic supralevator and the somatic infralevator pathways. These two pathways display proximal, intermediate and distal communications. The CN and branches from the DN communicate distally in an X-shaped region in which these pathways cross each other in two spatial planes. This interaction carries nitrergic activity to the distal part of the cavernous bodies in both sexes and is important for erectile function. The existence of these autonomic-somatic communicating branches suggests possible plasticity of the nervous supply of potential clinical interest for the prevention of postoperative urogenital and digestive dysfunctions and for pharmacological and rehabilitation treatments for these dysfunctions. However, these data were obtained in fetal tissues and further anatomical studies in adult are required, considering that anatomical relationship, as well nervous apparatus, could change anatomical relationships during puberty.