INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

PHARMACOLOGICAL STIMULATION of the paraventricular nucleus of the hypothalamus (PVN), through a variety of neuroactive compounds in conscious rats and electrical or pharmacological stimulation of the PVN in anesthetized rats, elicits penile erection and intracavernous pressure (ICP) increases (3, 8). The PVN contributes descending oxytocinergic fibers to the spinal cord (7), and the paraventriculospinal tract originates in the parvocellular part of the PVN (15). In male rats, the lumbosacral spinal cord contains oxytocinergic fibers (29), some of which synapse onto spinal preganglionic neurons (30). Furthermore, specific oxytocin (OT) binding sites are present in the sacral parasympathetic nucleus and the dorsal grey commissure of the L6-S1 spinal cord (34). Finally, PVN neurons are transsynaptically labeled with pseudorabies virus (PRV) injected in the corpus cavernosum (18). Engorgement of the penis with blood, leading to erection, is caused by increased blood flow to the penis and active relaxation of the erectile tissue of the corpora cavernosa and the corpus spongiosum (1). Both mechanisms are controlled by the autonomic nervous system. In rats, the sympathetic outflow to the penis originates in the T12-L2 spinal cord, and the proerectile parasympathetic outflow originates in the L6-S1 spinal cord (9). The sacral parasympathetic nucleus (SPN) of the L6-S1 spinal cord contains the preganglionic neurons that innervate the pelvic organs, including the penis. In a sexually relevant context, activation of proerectile neurons in the SPN may be elicited by information from peripheral or supraspinal origins (24). We tested the hypothesis that OT released by paraventriculospinal pathways could regulate the spinal control of penile erection through an effect on lumbosacral neurons in the rat.

	MATERIAL AND METHODS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Adult male Sprague-Dawley rats, sexually naïve and weighing 200-250 g, were purchased from Charles River (Saint-Aubin les Elbeufs, France). Rats were housed in groups of four in plastic cages containing wood-chip bedding. They had free access to commercial pelleted rodent chow (Piètrement, Provins, France) and tap water. Cages were placed in an animal facility maintained at 20°C and kept in a 12:12-h light-dark cycle (lights on at 8 AM). All animal experiments were carried out in accordance with the European Economical Community Directive of November 24, 1986 (86/609/EEC) on the use of laboratory animals. All efforts were made to minimize animal suffering and to reduce the number of animals used.

Experimental procedure. Rats were anesthetized with an intraperitoneal injection of urethane (1.2 g/kg in sterile water), and their temperature was maintained at 37°C using a homeothermic blanket. Intrathecal (it) catheterization was performed as reported by LoPachin et al. (17). Briefly, the rat's head was placed in a stereotaxic frame and was rotated nose downwards to facilitate catheter insertion. The catheter, a polyethylene tubing (PE-10) stretched to 150% of its original length in hot water, was cut to the required length so that its distal opening reached the L4-L6 or T12-T13 levels of the spinal cord. The skin and neck muscles were incised and retracted. The atlantooccipital membrane was opened, and the catheter, flushed with sterile NaCl 0.9%, was carefully advanced in the caudal direction. Finally, the rostral free end of the catheter was secured with the ligatures that closed the neck muscles and skin layers. The catheter was connected to a Hamilton syringe filled with saline to prevent cerebrospinal fluid leakage. Rats were tracheotomized to prevent aspiration of saliva and to perform artificial ventilation when muscle relaxant was used. The carotid artery and jugular vein were catheterized with polyethylene tubings filled with heparinized saline (25 U/ml) to record blood pressure (BP) via a pressure transducer (Elcomatic 750, Glasgow, UK) and inject drugs intravenously, respectively. ICP recording was performed as described previously (12). Tables 1 and 3 display the different groups of rats that we used. For it injections, compounds dissolved in 10 µl of NaCl 0.9% were injected within 10-20 s, immediately followed by a flush of 10 µl NaCl 0.9%. When we used cumulative injections of drugs, two consecutive injections were separated by a 15-min period. To perform pelvic nerve section (PNx rats), a suprapubic incision was performed. The pelvic nerves were exposed at the lateral aspect of the prostate, and nerves were sectioned 2-3 mm proximal to the major pelvic ganglion.

Data analysis. ICP rises were detected as increases in ICP pressure above a threshold. This threshold was calculated as the ICP value averaged over the 15-min period of observation before drug injection plus two standard deviations. The number of rats that displayed at least one ICP rise during the experiment, the number of ICP rises per rat averaged for each group, the number of ICP rises occurring within each period of 15 min separating two consecutive injections, the latency of the first ICP, and the duration of ICP rises were calculated. There exists a strong positive correlation between the amplitude of ICP increase elicited by cavernous nerve stimulation and BP (12). Therefore, in the present experiment, the amplitude of each ICP increase was reported to the corresponding diastolic BP (the ICP/BP ratio) (13). If several ICP rises occurred within the 15-min period separating two injections, then we averaged the ICP/BP ratio over the number of ICP rises. The area under the curve (AUC), also relative to diastolic BP and integrated over each 15-min period, was calculated. Where appropriate, results are expressed as mean values ± SE. Variables were evaluated by ANOVA followed by multiple-comparisons tests. Differences were considered statistically significant at P < 0.05.

Drugs. Urethane, OT, the OT agonist [Thr⁴,Gly⁷]OT, [Arg⁸]vasopressin (AVP), and hexamethonium (HXM) were purchased from Sigma (Saint-Quentin-Fallavier, France) and dissolved to the required concentration in NaCl 0.9%. Gallamine triethiodide (Flaxedil) was purchased from Specia (Rhône-Poulenc Rorer, Paris, France). The OT antagonist [(S)PMP¹,D-Trp²,Pen⁶,Arg⁸]OT was a generous gift from Dr. G. Flouret (Northwestern Univ., Chicago, IL).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

Cumulative injections of OT. Injection (it) of OT elicited ICP rises (Fig. 1A). Table 2 displays the number of rats, named responders, that had at least one ICP rise during the recording. Kruskal-Wallis one-way ANOVA on ranks demonstrated that there was a treatment effect on the number of responders (H = 41.8, df = 9, P = 3.6 × 10⁶). At least one ICP rise was recorded after injection of NaCl, OT at the L4-L6 and at the T12-T3 levels, OT iv, OT agonist (Fig. 1B), and OT at the L4-L6 level after curarization. In contrast, the groups that received injection of the OT antagonist followed by OT at the L4-L6 level, AVP, OT at the L4-L6 level after PNx, or HXM (P < 0.05 for each) included significantly fewer or no responders. Therefore, the effects of OT were impaired by the OT antagonist (Fig. 1C) by lesioning preganglionic fibers conveyed by the pelvic nerve and by blocking the transmission between pre- and postganglionic neurons. AVP had no effect on ICP.

View larger version (58K): [in this window] [in a new window]   Fig. 1.   Three recordings of blood pressure (BP; top traces) and intracavernous pressure (ICP; bottom traces; in mmHg) in 3 different anesthetized male rats. A: intrathecal injection of 0.9% NaCl (at t = 0 min), then each 15 min, increasing doses of oxytocin (OT; 0.3, 1, 3, 10, 30, 100, and 300 ng) were injected by the same route. B: same as A, but injection of the OT agonist [Thr4,Gly7]OT was performed. C: at t = 15 min, injection of the OT antagonist [(S)PMP1,D-Trp2,Pen6,Arg8]OT was performed, then same doses of OT as in A.

6

6

4

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Effects of OT (filled bars) and the OT agonist [Thr4,Gly7]OT (open bars) delivered intrathecally, cumulative doses (0-300 ng), on the number of ICP rises recorded in anesthetized male rats. OT (10 and 30 ng) elicited more ICP rises than the other doses (P < 0.05; 1). OT agonist (100 and 300 ng) elicited more ICP rises than the other doses (P < 0.05; 2). OT (10 ng) elicited more ICP rises than 10 ng of the OT agonist (P < 0.05; 3). OT agonist (300 ng) elicited more ICP rises than 300 ng of OT (P < 0.05; 4).

7

6

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Effects of OT (filled bars) and the OT agonist [Thr4,Gly7]OT (open bars) delivered intrathecally, cumulative doses (0-300 ng), on the ICP/BP ratio. OT (10, 30, and 100 ng) elicited the greatest ICP rises as measured by the ratio ICP/BP (P < 0.05; 1). OT agonist (100 and 300 ng) elicited the greatest ICP rises (P < 0.05; 2).

5

6

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effects of OT (filled bars) and the OT agonist [Thr4,Gly7]OT (open bars) delivered intrathecally, cumulative doses (0-300 ng), on the area under the curve (AUC) of ICP rises expressed by the AUC/BP ratio (arbitrary units). OT (10 and 30 ng) elicited the greatest AUC/BP (P < 0.05; 1). OT agonist (100 and 300 ng) elicited the greatest AUC/BP (P < 0.05; 2).

Injection of a single dose of OT. In a second group of experiments, we tested whether single injections of OT elicited ICP rises, based on a 30-min period of observation after injection (Fig. 5, Table 3). There was a significant dose effect of OT on the number of responders (H = 8.77, df = 3, P = 0.0326). The group that received 300 ng OT included significantly less responders (2 of 8) than the other groups (Table 3, 2nd column). Conversely, there was no dose effect of OT on the number of ICP rises [F(3,33) = 2.08, P = 0.1240]. Neither latency to the first ICP rise nor duration of ICP rises was dependent on the dose injected {[F(3,21) = 0.471, P = 0.7062] and [F(3,21) = 2.24, P = 0.1184, respectively]}. Finally, neither the ICP/BP nor the AUC/BP ratio was dependent on the doses injected (H = 3.73, df = 3, P = 0.292 and H = 7.64, df = 3, P = 0.0540, respectively).

View larger version (44K): [in this window] [in a new window]   Fig. 5.   Recording of BP (top trace) and ICP (bottom trace; in mmHg) in 1 anesthetized male rat after the intrathecal injection of 0.9% NaCl (at t = 0 min), then a single dose of 30 ng OT (arrow).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIAL AND METHODS RESULTS DISCUSSION REFERENCES

Results of the present study provide evidence for a proerectile effect of OT at the lumbosacral level in anesthetized rats. Delivering OT intravenously had no effect on ICP. OT receptors are present in the reproductive tract of males (11). OT contracts smooth muscles of the genital tract in vivo (20) and contracts the corpus cavernosum in vitro (32). Therefore, one cannot expect a proerectile (relaxant) effect of OT through an effect on a peripheral target.

Delivering OT at the T12-T13 level had no reliable effects on ICP. The thoracolumbar spinal cord contains sympathetic neurons destined to the penis (18). These sympathetic pathways are classically considered antierectile. Therefore, even if OT activated these neurons, the consequence could not be an erection. Proerectile effects should therefore be attributed to a specific targeting of OT on the lumbosacral spinal cord.

This hypothesis was confirmed by the fact that only rats that received OT at the lumbosacral level displayed the greatest number of ICP rises. We performed additional experiments to demonstrate that proerectile effects were OT specific. Only rats that received OT or its specific agonist, [Thr⁴,Gly⁷]OT, reliably displayed ICP rises. Vasopressin did not elicit such a response. Therefore, ICP rises were elicited by OT and not vasopressin. The presence of ICP rises occurring on it NaCl are likely spontaneous events. Their number is not constant between rats or the rank of injection. However, it suggests that erection may occur spontaneously under anesthesia in rats. After injections of the OT antagonist [(S)PMP¹,D-Trp²,Pen⁶,Arg⁸]OT, followed by OT, ICP rises were only recorded in one rat. It is unclear whether the ICP rises in this animal could be considered spontaneous. In this rat, we numbered two and three ICP rises in response to 100 and 300 ng OT, respectively. We hypothesize that in this animal, the dose of OT antagonist was not great enough to prevent responses to high doses of OT. We also consider that in the one rat that displayed ICP rises after vasopressin injection, such ICP rises occurred spontaneously. Indeed, they were recorded after an injection of 0.3 ng vasopressin. When compared with other groups, neither 0.3 ng OT nor its agonist elicited ICP rises.

After it injections of 30 ng OT, the number of ICP rises, their amplitude, and area were greater than those recorded with the other OT doses. After it injections of the OT agonist, 100 and 300 ng elicited the greatest numbers of ICP rises, and these reached the greatest amplitudes. The bell-shaped curve was evident when using OT, and rare or no ICP rises were recorded after 300 ng OT. In contrast, 300 ng of the OT agonist still elicited many responses. OT and [Thr⁴,Gly⁷]OT display the same affinity for the OT receptor present in uterine smooth muscle (2); a difference between spinal and peripheral OT receptors could explain why 300 ng OT does not elicit any erectile response in our experiment. It is unlikely that the saturability of the OT receptor accounts for the lack of ICP rises on 300 ng OT, because this dose should elicit at least as many ICP rises as 100 ng OT. It is interesting to note that 300 ng OT elicited no more ICP rises when injected as either cumulative or single doses. Could the desensitization of the OT receptor explain this decrease of ICP response? In cultured astrocytes, OT applications elicited calcium rises in which amplitude decreased if the application was repeated. The authors observed a 20-min wash period between two applications before they could record a full recovery of the calcium response (10). In our experiment, a period of 15 min separated two consecutive injections, which suggests that no recovery could occur in this condition. OT (10, 30, and 100 ng) elicited fewer ICPs on single doses compared with cumulative doses. Furthermore, the ICP/BP ratio displayed a bell-shaped response curve for cumulative treatments, but no such profile was noted on single-dose treatments. It remains unclear to us whether such differences rely on time of exposure of the receptor to OT or to an interaction between time of exposure and dose.

At high doses, OT may bind vasopressin receptors. The latter are present in the lumbosacral spinal cord of rats (33). Although 10 times less potent than vasopressin, OT could act on the V₁ receptor of sympathetic preganglionic neurones of the neonate rat spinal cord (27). In our experiment, OT acting at vasopressin receptors could display inhibitory effects on spinal proerectile neurons. We tested the effects of it vasopressin. The peptide had no effect on ICP. In contrast, it could oppose a proerectile effect of OT.

SPN neurons convey parasympathetic fibers to the penis through the pelvic and cavernous nerves (14, 18, 21). Electrical stimulation of the pelvic and cavernous nerves elicits ICP rises in anesthetized rats (12, 13, 28). By delivering HXM iv, we blocked nicotinic receptors, thereby inhibiting the synaptic transmission between preganglionic fibers of the pelvic nerve and postganglionic fibers in the cavernous nerve. After HXM injection, no ICP rise occurred in response to it OT, demonstrating that OT recruited preganglionic neurons. The bilateral section of the pelvic nerve also prevented any ICP increase to occur after OT injection. Therefore, in the present study, the proerectile effects of OT are caused by activation of the sacral parasympathetic outflow. However, oxytocinergic activation of sacral parasympathetic pathways may be direct or relayed through interneurons present in the dorsal grey commissure of the lumbosacral spinal cord. Tight anatomic relationships between the two areas in the rat spinal cord, as demonstrated with transsynaptic transport of PRV from the corpus cavernosum, confirm this hypothesis (18). Furthermore, both areas contain OT receptors (34). Therefore, OT released by descending PVN-spinal pathways may activate both interneurons and preganglionic neurons.

Penile erection in conscious mammals recruits autonomic pathways to the penis and somatic pathways to the perineal striated muscles (25). In conscious mammals, contraction of striated muscles on the erect penis elicits peaks of penile pressure rises that largely override BP (5, 23, 26). According to some authors, bulbospongiosus (BS) and ischiocavernosus (IC) motoneurons receive descending projections from the PVN (35). Furthermore, transsynaptic retrograde labeling from the IC or BS muscles using PRV or rabies virus labels some neurons in the PVN (19, 31). According to these data, OT could also control the somatic outflow to the perineal striated muscles. However, although rare OT-immunoreactive fibers have been demonstrated in the ventral horn of the rat spinal cord (30, 34), this area does not contain OT receptors (34). After OT injection, we never recorded ICP rises over BP, and the injection of the striated muscle blocking agent gallamine triethiodide did not affect the ICP rises elicited by OT. Therefore, our data demonstrate an effect of OT on penile pressure, independent of striated muscle, and suggest a lack of excitatory effect of OT onto IC and BS motoneurons.

OT may activate lumbosacral parasympathetic neurons and interneurons destined to pelvic organs other than the penis. Indeed, it was demonstrated that in conscious female rats, it OT increased micturition pressure and decreased bladder capacity and micturition volume (22). Interestingly, the most efficient dose in this model was 30 ng OT. We also identified 30 ng as the dose of OT that yielded the greatest probability of eliciting ICP rises when injected in cumulative doses, and only the number of responders when OT was injected as single doses. If comparable doses of OT activate different parasympathetic outflows, then it remains to be determined how the spinal network integrates this increase of OT, because all pelvic viscera are not active at the same time. It may be suggested that in a sexually relevant context, it is the convergence of information from the periphery and from supraspinal structures that elicits the specific activation of proerectile pathways at the spinal cord level.

In conscious rats, noncontact erections and drug-induced erections (e.g., apomorphine-induced erections) reflect the activity of supraspinal nuclei. These erections are transient and repetitive, and recording ICP during noncontact and apomorphine-induced erections revealed transient rises of ICP (6). In our experiments, it injections of OT elicited phasic ICP rises. It suggests that the spinal cord translates the tonic excitation by supraspinal nuclei or by OT into the phasic activation of parasympathetic pathways leading to phasic ICP rises.

In our experiment, OT would be a potent activator of the spinal generator of penile erection. Once this rhythmic generator is activated, OT could not further regulate the number, the amplitude, and the area of the ICP rises, as evidenced by lack of dose effect of OT injection on the ICP/BP and AUC/BP ratios.

Pharmacological stimulation of the PVN in conscious rats and its electrical stimulation in anesthetized rats elicit penile erection (4, 8). Lesions of the PVN suppress apomorphine-induced erections (4) and impair noncontact erections (16). In rats, fibers issued from the parvocellular part of the PVN reach the lumbosacral spinal cord (7). The SPN contains OT-immunoreactive fibers and OT receptors (30, 34). Our results suggest that by delivering OT at the lumbosacral level, we mimicked the release of OT by PVN-spinal fibers in rats. The present results demonstrate that OT exerts proerectile effects, as measured through increases in ICP, when it is delivered at the L4-L6 spinal cord in anesthetized rats. They demonstrate that the effects are specific, being mimicked by a specific agonist but not by arginine-vasopressin, and are blocked by a specific OT antagonist. Proerectile effects of OT are due to the activation of autonomic efferent pathways running in the pelvic nerves.

Perspectives