- Email: [email protected]
Reflex penile erection in anesthetized mice: An exploratory study
Neuroscience 155 (2008) 283–290
REFLEX PENILE ERECTION IN ANESTHETIZED MICE: AN EXPLORATORY STUDY J. ALLARD* AND N. J. EDMUNDS
eral input from the pelvis (see McKenna, 2000 and Allard et al., 2005, for review). Thus, from a physiological and mechanistic endpoint, penile erection and ejaculation can be viewed as reflex responses modulated by the brain. The afferent limb of this reflex loop consists of sensory inputs from the penis, which are subsequently conveyed in their distal segment by the dorsal penile nerve (DPN) followed by the sensory branch of the pudendal nerve. The efferent limb is more complex, comprising parasympathetic, sympathetic and somatic projections, which are involved in penile erection, the emission and the expulsion phase of ejaculation respectively. Proerectile parasympathetic fibers reach the penis through the cavernous nerve. Sympathetic fibers, that contribute to the emission phase, join the sexual accessory glands through the hypogastric nerve. Finally, the rhythmic contractions of the bulbospongiosus (BS) muscles which are responsible for the expulsion phase of ejaculation, are controlled by motoneurons that project through the motor branch of the pudendal nerve. Most of our knowledge of the pharmacology and neurophysiology of the male sexual reflex has been generated in experiments using male rats. In the urethane-anesthetized rat, McKenna et al. (1991) showed that peripheral stimulation of genital origin, through mechanical stimulation of the penile glans and urethra, could result in penile erection and rhythmic contractions of the BS muscle (see also Carro-Juarez and Rodriguez-Manzo, 2000, 2005; Carro-Juarez et al., 2003). These contractions of the BS muscle are considered a surrogate marker of the expulsion phase of ejaculation. A variation of this original model replaced the mechanical stimuli of the genital tract by direct electrical stimulation (ES) of the penile sensory afferent nerves running in the DPN (Pescatori et al., 1993; Allard et al., in press). A key feature of these experimental models is that complete spinal transection at the thoracic level is necessary for penile erection and the expulsion reflex to be elicited. The likely reason for this is that in intact rats, the spinal network controlling sexual reflexes is tonically inhibited by descending input from the brainstem (Marson and McKenna, 1990). The requirement for spinal transection in this model gives a unique opportunity to specifically interrogate the spinal circuitry controlling sexual reflex. However, from a preclinical research perspective spinal transection precludes the ability to investigate mechanisms that act via central modulation of descending inhibitory input. For example chronic treatment with selective 5-HT reuptake inhibitor modulates descending input causing an increase in ejaculatory latencies in patients suffering from depression or premature ejaculation (see
Pfizer Global Research and Development, Genitourinary TA, Ramsgate Road, Sandwich, Kent, CT13 9NJ, UK
Abstract—Ejaculatory-like rhythmic contractions of the bulbospongiosus (BS) muscle and penile erection can be elicited in the urethane-anesthetized rat, following spinal cord transection, upon electrical stimulation (ES) of the dorsal penile nerve (DPN). The aim of this work was to investigate this reflex in anesthetized mice. Adult C57BL6 mice were anesthetized with isoflurane. The BS muscle and corpus cavernosum were instrumented to allow quantification of the BS muscle electromyographic activity (BS EMG) and intracavernosal pressure respectively. The femoral artery and jugular vein were catheterized to allow measurement of blood pressure and compound administration. ES of the DPN, via bipolar silver electrodes, reliably evoked erectile responses in mice with intact spinal cords. The overall amplitude of the erectile response was frequencyand pulse duration– dependent. Erectile responses were abolished by bilateral cut of the sensory branch of the pudendal nerve. Transection of the spinal cord potentiated the erectile responses and increased the area under the curve of the BS EMG when compared with those animals with intact spinal cords. However, no coordinated rhythmic contractions of the BS muscle during or after the ES could be observed, with or without spinal transection. Melanotan-II failed to enhance the erectile response induced by ES of the DPN, in mice with intact spinal cords. ES of the DPN in isoflurane-anesthetized mice could be a useful model in which to study the interplay between brain and spinal cord in the control of reflex penile erection, and could take advantage of knockout mice models. However, the lack of efficacy of Melanotan-II suggests that further experiments are necessary to confirm the future utility of this model. In contrast to rats, the expulsion reflex could not be reliably elicited in mice with or without spinal transection. This latter finding suggests the existence of fundamental differences in the organization of the spinal network controlling sexual reflexes between rats and mice. © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: dorsal penile nerve, bulbospongiosus muscle, ejaculation, Melanotan-II.
Penile erection and ejaculation are controlled by spinal autonomic centers, the activity of which is dependent upon descending input from supraspinal structures and periph*Corresponding author. Tel: ⫹44-1304646005; fax: ⫹44-1304651987. E-mail address: [email protected] (J. Allard). Abbreviations: AUC, area under the curve; BP, blood pressure; BS, bulbospongiosus; DPN, dorsal penile nerve; EMG, electromyographic; ES, electrical stimulation; ICP, intracavernosal pressure; ICPmax, maximal intracavernosal pressure; ICPmean, mean intracavernosal pressure; MT-II, Melanotan-II.
0306-4522/08$32.00⫹0.00 © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2008.05.027
283
284
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
Table 1. Overview of the ICPmax/BP, ICPmean/BP and ICPAUC/BP of the erectile responses generated with similar or identical ES consisting of square pulses (6 V, 1 ms) at 3 Hz for 30 s Group #
N
n
ICPmax/BP
ICPmean/BP
ICPAUC/BP (s)
BP (mmHg)
1 2 4 5 6
3/4 4/4 3/4 3/4 3/4
20 12 2 2 3
0.70⫾0.02 0.70⫾0.05 0.77⫾0.01 0.66⫾0.06 0.67⫾0.03
0.42⫾0.01 0.55⫾0.04 0.49⫾0.06 0.41⫾0.12 0.43⫾0.03
14.2⫾0.4 35.4⫾7.7 22.9⫾0.6 16.0⫾4.7 21.3⫾3.0
83.8⫾5.7 65.1⫾2.3 83.0⫾2.5 79.0⫾2.6 75.7⫾3.8
(SCintact) (SCtransected) (Pulse duration) (Frequency)a (MT-II test)
a
The erectile responses mentioned in the frequency group were generated with ES at 4 Hz. Group 3 (which used three mice with intact and cut spinal cord) was omitted from the comparison because the duration of the train of stimulation was consistently shorter (20 s). “N ” corresponds to the number of mice included for quantification/the number of mice originally included in the group. “n” corresponds to the number of responses analyzed per mice.
(Hirschfeld, 2003; Waldinger, 2007) for review). Since reflex contractions of the BS muscle have been reported following urethral distension in urethane-anesthetized mice with intact neuraxis (Burnett et al., 1998), we hypothesized that the sexual reflex, initiated through stimulation of the DPN, could be generated in anesthetized mice with intact spinal cords. The present study aimed to characterize the reflex sexual response of anesthetized mice to ES of the DPN.
EXPERIMENTAL PROCEDURES Surgical preparation C57BL/6 mice (Charles-River, Manston, UK) were anesthetized with isoflurane (4% for induction, 2.5% for surgery, 1.8 –2.0% for recording) with level of anesthesia being assessed by lack of withdrawal reflex upon pinching of the paw. Body temperature was maintained at 37 °C using a homeothermic blanket. The femoral artery was catheterized with a stretched catheter (approximate external diameter, 0.1 mm) for blood pressure (BP) recording. The jugular vein was additionally catheterized for drug delivery when necessary. The skin surrounding the penis was removed and the connective tissue surrounding the right corpus cavernosus carefully removed. A dorsal bundle including the DPNs and dorsal penile vein and arteries was dissected from the underlying corpus spongiosus (it was technically difficult to isolate the DPN alone). A midline incision was made on the scrotal skin and the right BS muscle exposed. A 26 gauge needle connected to a pressure transducer was inserted in the right corpus cavernosus to record intracavernosal pressure (ICP). Both the BP and ICP catheter were filled with heparinized saline (50 IU/ml). Electromyographic (EMG) activity of the BS muscle was measured with bipolar differential recording. The electrode was made from a pair of 26 gauge needles. The tip of the needle was bent to form a hook shape that allowed the needle to stay in place once inserted in the right BS muscle. A bipolar stimulating electrode made of 0.2 mm diameter silver wire was hooked around the dorsal bundle. ESs were performed using a Grass stimulator (Grass S-88, Grass Instruments, Quincy, MA, USA). For spinal transection, an incision was made over the midline on the back at the thoracic level. The membrane connecting the T7 and T8 vertebrae was removed with bone rongeur and the spinal cord transected with spring scissors. Muscle and skin layers were closed separately with staples. For comparison, recordings were also generated with Sprague–Dawley rats (250 –350 g, Charles-River) in almost identical conditions, the difference being that the DPN was bilaterally dissected from the surrounding tissue before being hooked on the stimulating electrodes. In rats or mice, ES of the DPN started 30 min after transection of the spinal cord with the entire protocol lasting a further 30 – 60 min.
Experimental groups Mice were divided into six different experimental groups (Table 1). The reproducibility of successive reflex responses to identical ES (train of square wave pulses of 6 V and 1 ms duration at 3 Hz for 30 s) in mice with spinal cord intact and cut was assessed in group 1 and 2 respectively. The potentiation of reflex responses by spinal transection was assessed in group 3 with ES consisting of trains of square wave pulses of 6 V and 0.1 ms duration at 1, 3 and 9 Hz for 20 s, using mice with spinal cord intact and cut. Preliminary data suggested that such parameters of stimulation would span a range from submaximal to supramaximal erectile responses. A similar experiment was performed in rats using stimulation parameters appropriate for this species. The pulse-duration and frequency dependency of the reflex response were assessed in groups 4 and 5, using train of square wave pulses of 6 V and 0.1–5 ms duration at 3 Hz for 30 s (group 4) or 6 V and 1 or 0.1 ms duration at 0.5– 8 Hz for 30 s (groups 5a and b respectively). ES were performed every 3 min. When different pulsedurations or frequencies were tested, each ES was repeated twice. The potential proerectile activity of Melanotan-II (MT-II, Bachem UK Ltd., Merseyside, UK) on the erectile responses induced by ES of the DPN in mice with spinal cord intact was assessed in group 6. Train of square wave pulses of 6 V and 1 ms duration at 3 Hz for 30 s were performed every 3 min until two consecutive equivalent responses were obtained. Three ES at 1 Hz were then performed (pre-injection ES). MT-II (1 mg/kg) or the corresponding vehicle (saline, 1 ml/kg) was then injected i.v. 30 s after the last ES at 1 Hz. The injection was performed over 30 s. Three ES at 1 Hz were performed again every 3 min, starting 3 min after completion of the i.v. injection. For each mouse, the mean maximal intracavernosal pressure (ICPmax)/BP, mean intracavernosal pressure (ICPmean)/BP and ICPAUC/BP of the three erectile responses after the drug or vehicle injection were expressed as a percentage of the mean before the injection.
Recording analysis The analog signal from the pressure transducer (DTX Plus, Becton Dickinson, Milton Keynes, UK) was amplified using a NL108 amplifier (Digitimer Ltd., Hertfordshire, UK). The EMG signal was amplified and filtered (10 –50 Hz low, 10 –50 kHz high) using NL125 and NL126 module respectively (Digitimer Ltd.). Filters and amplifiers were set into a 900 D Neurolog box (NL905, Digitimer Ltd.). Analog signals were digitized using a DT-P6 converter (Digitimer Ltd.), and fed to a PC. Pressure and EMG signals were sampled at 500 and 5000 Hz respectively. Acquisition and analysis was performed with notochord-hem 4.1.0.38. Erectile responses were quantified using the ratio of the maximal and mean value of the ICP during the response over the corresponding mean BP during the response (ICPmax/BP and ICPmean/BP respectively), and the ratio of the area under the curve (AUC) of the ICP during the response over the correspond-
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
285
Fig. 1. Upper and lower traces correspond to BP and ICP recordings, respectively, in a mouse with an intact spinal cord. ES (square wave pulses of 6 V, 1 ms at 5 Hz for 15 s, bar in abscissa) of the DPN gave rise to immediate erectile response (arrow) that were abolished (arrowhead) after bilateral transection of the sensory branch of the pudendal nerve (PNsbx). Note the change in ICP baseline activity after PNsbx.
ing mean BP (ICPAUC/BP, expressed in s). Since the magnitude of a given erectile response is dependent on BP, all ICP measures were divided by the corresponding BP in order to normalize ICP responses to systemic BP. The beginning and the end of each erectile event were determined manually. ICPmean, ICPmax and AUC were computed from a baseline determined by the mean value of the ICP during the 5 s preceding the cursor delimiting the beginning of the erectile response. The AUC of the rectified BS EMG was computed from the end of the ES (to avoid integrating the stimulation artifact) until the return of BS muscle activity to baseline. For a given parameter, the ratio of the standard deviation over the mean (coefficient of variation) was computed when necessary to evaluate the reproducibility of the reflex response. Data are expressed as mean⫾standard error of the mean.
Statistical analysis Data per group correspond to the mean of each individual mean per parameter. Experimental groups 1, 2, 4 and 5 were generated successively out of an experimental plan and no statistical comparison was performed between and within these groups. In contrast, experiments corresponding to groups 3 and 6 were designed to allow statistical analysis. For these groups, comparison of the
relevant parameters was performed using Student’s t-test. Differences were considered significant when P⬍0.05. All experiments were conducted in accordance with the European Community Council Directive (86/609/EEC), with the UK Animals (Scientific Procedures) Act 1986 and subject to local ethical review. All efforts were made to minimize the number of animals used and their suffering.
RESULTS Exploratory work showed that ES of the DPN resulted in erectile responses in intact mice anesthetized with isoflurane (data not shown). To assess whether the erectile responses induced by ES of the DPN resulted from a spinally mediated reflex response, ES (square wave pulses of 6 V, 1 ms at 5 Hz for 15 s) of the DPN were performed before and after bilateral transection of the sensory branch of the pudendal nerve within the ischiorectal fossa in two mice. Transection of the sensory branch of the pudendal nerve completely suppressed the erectile response to ES of the DPN while increasing baseline ICP
Fig. 2. Upper and lower traces correspond to BP and ICP recordings, respectively. Recordings in A correspond to a mouse with an intact spinal cord, and show the erectile responses to 20 successive ES of the DPN. Erectile responses progressively reach a plateau level and are stable thereafter. Recordings in B correspond to a mouse with the spinal cord transected at the thoracic level, and show the erectile responses to 12 successive ES of the DPN. Note the considerable variation of the baseline ICP and the prolonged duration of the erectile response.
286
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
activity (see Fig. 1 for example of recording). Further exploratory work showed that ES, consisting of a train of square wave pulses of 6 V and 1 ms duration at 3 Hz for 30 s, resulted in maximal or close to maximal erectile responses in three mice (data not shown). These stimulation parameters, which were derived from work in rats in a similar model, were therefore used as a starting point to study the erectile responses to ES of the DPN in isoflurane-anesthetized mice. Groups 1 and 2: reproducibility of successive erectile responses The reproducibility and stability of the erectile responses induced by ES of the DPN were studied in a group of four mice (group 1). One mouse failed to respond to ES of the DPN and was excluded from the quantification. In the remaining three mice, at least 20 consecutive erectile responses could be elicited (Fig. 2A). Analysis of the recordings showed that four to six consecutive ES were necessary for the erectile responses to stabilize (Fig. 2A and Fig. 3). Erectile responses remained relatively stable thereafter, although a slight progressive decrease could be observed after approximately 15 stimulations (Fig. 3). When considering all erectile responses for each individual mouse, the coefficient of variation for the ICPmax/BP never exceeded 14%, making it the most stable parameter characterizing the erectile responses (Fig. 3A, Table 1). ICPmean/BP and ICPAUC/BP were more variable, with coefficient of variation reaching 26% and 28% respectively (Fig. 3B and C, Table 1). The AUC of the rectified BS EMG response (mean value in the group, 22.5⫾3.9 uV·s) was quite variable between stimulations within each mouse, with a coefficient of variation ranging from 29% to 66%. No rhythmic, coordinated contractions of the BS muscle, similar to an ejaculatory pattern, could be detected during or at completion of the ES. Because sexual reflexes are tightly controlled by descending projections from the brain to the lumbo-sacral spinal cord, the reproducibility and stability of the reflex responses induced by ES of the DPN were studied in four mice with complete spinal cord transection at the thoracic level (group 2). Note that since groups 1 and 2 were experimentally and temporally independent of one another, no statistical comparison was attempted between these groups. BP was consistently lower in mice with spinal cord cut compared with spinal cord intact (Table 1). The baseline ICP was noticeably variable at the beginning of the recording in mice with complete spinal cord transection (Fig. 2B). The fluctuation of the baseline tended to decrease throughout the recording, but was still present in three of four of the mice studied at completion of the recording. Following transection at least 12 consecutive erectile responses could be elicited (although more ES could have been performed). For each mouse, the coefficient variation for the ICPmax/BP and ICPmean/BP never exceeded 9% and 17% respectively (Fig. 3A–B, Table 1). In contrast, the ICPAUC/BP was more variable, with a coefficient of variation reaching 38% in two mice (Fig. 3C, Table 1). The AUC of the rectified BS EMG was highly
Fig. 3. Successive erectile responses generated with ES at 3 Hz (pulses at 6 V, 1 ms) for 30 s were quantified in mice with spinal cord intact (filled symbol, 20 responses, three mice) or cut (open symbol, 12 responses, three mice). The maximal and mean value of the erectile response, and the AUC, corrected by the BP, are displayed in A (ICPmax/BP), B (ICPmean/BP) and C (ICPAUC/BP) respectively. The AUC of the BS EMG is displayed in D. The abscissa corresponds to the successive erectile responses. Mice with spinal cord cut displayed larger ICPAUC/BP, and more variable EMG responses of the BS muscle.
variable both between mice (mean value in the group, 98.3⫾73.4 uV·s) and within mice (coefficient of variation ranging from 25% to 89%). No coordinated rhythmic contractions similar to that occurring during an ejaculatory response could be detected during or at completion of the ES in mice (see Fig. 4 for example of BS EMG recording in mice and rat). Group 3: impact of spinal transection on erectile responses Comparison of the erectile responses generated with ES at 3 Hz showed that the ICPmax/BP and ICPmean/BP were similar in mice with spinal cords either transected or intact (Fig. 3A and B, Table 1). In contrast, the ICPAUC/BP of the erectile response was larger in mice with transected spinal cord compared with intact spinal cord (35.4⫾7.7 s and 14.2⫾0.4 s respectively), because of a longer duration of the erectile response. This suggests a suppressive spinal influence on the erectile response to ES of the DNP, which was also revealed by the shorter latency (time between the start of the ES and the rise of the ICP) in mice with spinal cord cut (1.7⫾0.1 s) compared with spinal cord intact (5.7⫾0.4 s).
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
287
Fig. 4. The ICP and BS EMG recordings in response to ES of the DPN in a mouse with spinal cord cut are shown in A and B respectively. The recording in C is a magnification of the recording in B (magnified zone is delimited by the square in B). Note that the ICP rises within seconds after the onset of the ES (A). Magnification of the EMG recording in mice (C) does not show any sign of coordinated activity of the BS muscle. In contrast, magnification of the corresponding BS EMG response in rat with spinal cord cut (D) in identical experimental condition evidences reflex rhythmic coordinated (arrows) contractions of the BS muscle, similar to what is observed during the expulsion phase of ejaculation. In both animals, the response was triggered by ES at 3 Hz.
To further characterize the potentiating effect of spinal transection on the reflex erectile response, a head to head frequency-response comparison was made between intact and spinal cord transected mice (group 3, three mice in each condition). Electrical parameters generating submaximal erectile responses (train of square wave pulses of 6 V, 0.1 ms at 1, 3 and 9 Hz for 20 s) were used to unmask any potentiating effect of spinal transection. Erectile responses generated with ES at 1 Hz appeared dramatically increased in animals with spinal transection (Fig. 5A–C), although this did not prove statistically significant (ICPmax/BP, P⫽0.057; ICPmean/BP, P⫽0.059; ICPAUC/BP, P⫽0.063). However, for erectile responses generated at 9 Hz, the ICPmean/BP and ICPAUC/BP were significantly greater in mice with spinal cord cut than intact (P⫽0.022 and 0.013 respectively). As comparison, a similar experiment was performed in rats. Under isoflurane-anesthesia, the reflex erectile response-frequency curve is rightward shifted by almost one order of magnitude in rat with intact spinal cord compared with mice (data not shown). Therefore, reflex erectile responses were generated in rats with spinal cord intact and transected (four in each group) using train of square wave pulses of 6 V, 1 ms at 1, 3 and 30 Hz for an overall duration of 20 s to generate subthreshold to suprathreshold stimuli (Fig. 5D–F). Erectile responses generated at 1 and 3 Hz were of much greater amplitude in rats with spinal cord transected than intact (Fig. 5D–F). All parameters characterizing erectile responses were significantly increased in rat with spinal cord cut compared with intact with ES at 3 Hz (ICPmax/BP, P⬍0.001; ICPmean/
BP, P⫽0.020; ICPAUC/BP, P⫽0.018). Erectile responses generated with ES at higher frequencies (30 Hz) were comparable in rats with either intact or transected spinal cords. In addition, reflex rhythmic contractions of the BS muscle were observed in four of four rats with spinal cord transection upon ES at 3 and 30 Hz (see Fig. 4D for example of recording). Only one rat with spinal cord intact displayed such contractions upon ES at 30 Hz. Groups 4 and 5: pulse-duration and frequency dependency of the erectile response To further characterize the erectile response to ES of the DPN, the relationship between the duration of the pulse and the resulting erectile response was studied in a group of four mice (group 4). One mouse did not respond to ES of the DPN and was excluded from the quantification. The amplitude of the response increased with the pulse duration from 0.01– 0.5 ms and reached a plateau thereafter (see Fig. 6A for example of recording, and Fig. 6B and C for quantification). Then, the relationship between the frequency of the pulses and the resulting erectile response was studied with train of square wave pulses of 0.1 or 1 ms duration (corresponding to submaximal and maximal pulse duration respectively) in two independent groups of three and four mice respectively (groups 5a and b). One mouse in the 1 ms group failed to respond. All other mice gave consistent responses and were included in the quantification. For both pulse durations, there was a frequency-dependent increase in the amplitude of the erectile response
288
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
the erectile response (see Fig. 6D for example of recording with 0.1 ms duration pulse). Responses induced with 1 ms pulse were greater than responses induced with 0.1 ms pulse at 0.5 and 1 Hz (Fig. 6E and F). Beyond 2 Hz, there was no difference in the responses generated with pulses of 0.1 or 1 ms. The greatest difference was observed at 0.5 Hz, for which the ICPmax/BP was 0.04⫾0.01 and 0.2⫾0.1, and the corresponding ICPAUC/BP 0.7⫾0.2 and 3.1⫾1.2 s with 0.1 and 1 ms pulse respectively. Group 6: effect of MT-II on erectile responses
Fig. 5. ES of the DNP were performed in mice (A–C) or rats (D–F) with spinal cord intact (open bars) or transected (filled bars). ES consisted in square pulses of 6 V, 1 ms at 1, 3 and 30 Hz in rats, and square wave pulses of 6 V, 0.1 ms at 1, 3 and 9 Hz in mice (total duration 20 s for both). Low frequency of ES (1 Hz in mice, 1 and 3 Hz in rat) clearly unraveled the potentiating effect of spinal cord transection of the erectile response in mice and rat. ES at supramaximal frequency (30 Hz in rat, 9 Hz in mice) overcame the effect of spinal transection in rats, but not in mice. Statistical comparison (t-test) was made in each species between spinal cord cut and intact for each frequency (* P⬍0.05; ** P⬍0.01).
from 0.5–2 Hz (see Table 1 for results at 4 Hz). Increasing the frequency above 2 Hz did not increase the amplitude of
Table 1 summarizes the values of the parameters characterizing the erectile responses to ES at 6 V, 1 ms and 3 Hz in the different independent groups generated throughout the present study. Variability of the ICPAUC/BP response between the different groups of animals suggested that the effect of drug treatment would be better observed within individual mice than between different groups. Therefore, erectile responses were generated at 1 Hz before and after the i.v. injection of 1 mg/kg MT-II or the corresponding vehicle (saline) in two groups of five mice. Two mice in the vehicle group failed to respond to ES of the DPN and were excluded. The ICPmax/BP, ICPmean/BP and ICPAUC/BP of the erectile responses after the injection of MT-II (n⫽5) represented 93⫾5, 94⫾16 and 103⫾17% of the response before the injection, compared with 97⫾4, 93⫾8 and 104⫾12% after the corresponding vehicle injection (n⫽3, see Table 1 for basal responses). Thus, there was no proerectile facilitator effect of MT-II 1 mg/kg in the present model.
Fig. 6. Recording in A (BP and ICP corresponding to upper and lower trace respectively) illustrates the effect of increasing pulse duration (from 0.01–5 ms) on the erectile responses to ES (pulses of 6 V delivered at 3 Hz for 30 s) of the DPN in mice with spinal cord intact. Quantification of similar recordings in three mice shows the pulse-duration dependency of the ICPmax/BP and ICPmean/BP (square and triangle respectively, B) and the ICPAUC/BP (C) of the erectile response. Recording in D illustrates the effect of increasing frequency (from 0.5– 8 Hz) on the erectile responses to ES (pulses of 6 V, 1 ms for 30 s) of the DPN. Similar recordings were generated in two groups of three mice using pulses of 0.1 ms or 1 ms duration. Black and gray traces (1 and 0.1 ms respectively) illustrate the corresponding frequency dependency of the ICPmax/BP and ICPmean/BP (square and triangle respectively, E) and the ICPAUC/BP (F) of the erectile response.
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
DISCUSSION The exploratory work presented herein demonstrates that ES of the DPN induces frequency dependent reflex erectile responses in isoflurane-anesthetized C57BL/6 mice with intact spinal cords. These reflex erectile responses could be potentiated by complete transection of the spinal cord. One consideration regarding the present experiments concerns the lack of erectile response to ES of the DPN in 20% of the mice. We have no rational explanation for this finding at present, since there were no apparent differences in surgical preparation or depth of anesthesia between groups. Attempts to use other anesthetics (urethane 2 g/kg, ketamine/medetomidine 75 mg/kg/1 mg/kg and ␣-chloralose/ketamine 120 mg/kg/100 mg/kg delivered i.p., and propofol 13 mg/kg/min for 2 min followed by 2–3 mg/ kg/min delivered i.v.) were unsuccessful (data not shown). In our hand it was extremely difficult to obtain a stable plan of anesthesia with these anesthetics, and they did not reduce the number of non-responding mice. Thus, in our opinion, isoflurane is the most practical and reliable anesthetic to induce erectile response to ES of the DPN. There is no functional description of the sacral plexus and pudendal nerve anatomy in mice. The sensory branch of the pudendal nerve, which is the proximal end of the DPN, was therefore identified according to the description made in rat by Pacheco and co-workers (1997). Bilateral transection of the identified sensory branch of the pudendal nerve eliminated the erectile response to ES of the DPN. This suggests that the erectile responses generated in the present experiment involved a reflex loop comprising the DPN/sensory branch of the pudendal nerve (afferent arm) and the pelvic nerve/cavernous nerve (efferent arm). The increased ICP baseline activity that was observed after cutting the nerve might be due to transection-induced firing within the proximal end that would partially activate the reflex loop (R. D. Johnson, personal communication). Further experiments involving transection of the efferent arm (i.e. the pelvic and/or cavernous nerve) would be necessary to confirm the existence of this reflex erectile loop in mice, as has been previously established in rats (Rampin et al., 1994). The erectile responses to ES of the DPN were pulse duration-dependent for a set frequency. The pulse-duration dependency might correspond to a progressive increase in the number of fibers recruited by the ES and/or a change in the subtype of fibers (from A to A⫹C) recruited by the ES within the DPN. Concomitant recording of conduction velocity (neurogram) and erectile response for a given parameter of ES would be necessary to determine whether A fibers alone, or A and C fibers are involved in the generation of reflex erectile response in this model. The erectile responses were also frequency-dependent for a set square wave pulse voltage/duration. The presence of a graded response to increasing frequency of stimulation suggests that this model is suitable to assess ability of a drug to facilitate erections (see later), in the same way as ES of the cavernous nerve has been used to illustrate the proerectile activity of PDE5 inhibitors in mice or rats (see
289
for example Mizusawa et al., 2001; Giuliano et al., 2003; Champion et al., 2005). Interestingly, the frequency necessary to reach the plateau response upon ES of the DPN (2–3 Hz) was consistently lower than the frequency necessary to reach the plateau response upon ES of the cavernous nerve in similar conditions of anesthesia in the same mice strain (10 –15 Hz) (Behr-Roussel et al., 2006). This suggests a consistent amplification of the afferent signal, probably at the level of the spinal cord. Burnett et al. (1998) claimed that urethral distention induced reflex ejaculatory-like rhythmic contractions of the BS muscle (expulsion reflex) in urethane-anesthetized C57BL/6x129/SvEv mice with the spinal cord intact. In the present study, no expulsion reflex could be observed following ES of the DPN in isoflurane-anesthetized C57BL/6 mice with spinal cord intact. We believe that the present results are indeed in agreement with the report of Burnett and co-workers, as close examination of their data (see Fig. 2a) does not evidence the coordinated contractions of the BS muscle normally associated with ejaculation, in response to urethral stimulation. Nevertheless, an apparently greater non-coordinated activity of the BS muscle was observed following the stimulation in the study of Burnett and coworkers, which was not observed herein. This difference might be explained by the difference in strain or anesthetic used, or the type of stimulation. In agreement with the general view that descending projections from the brainstem exert a tonic inhibition on the spinal structures controlling reflex erection (McKenna, 2000), spinal cord transection potentiated the erectile response to ES of the DPN in mice. Nevertheless, the present study suggests differences in the spinal network controlling reflexive erections between rats and mice. In animals with spinal transection, the latency of the ICP rise in mice was much shorter when compared with rats (within seconds of the start of the ES and at completion of the ES respectively). Furthermore, in contrast to urethane-anesthetized rats (Marson and McKenna, 1990; McKenna et al., 1991; Pescatori et al., 1993), spinal cord transection did not allow the generation of rhythmic contractions of the BS muscle in response to ES of the DPN in isoflurane-anesthetized mice (present study). This difference cannot be solely attributed to the properties of isoflurane, as we have shown here that in isoflurane-anesthetized rats with spinal transection, ES of the DPN results in rhythmic contractions of the BS muscle. The differences observed herein between mice and rat sexual responses could well be a reflection of differences in their sexual behavior. During sexual behavior, rats require several brief intromissions to reach ejaculation, whereas mice require prolonged intromission with repeated thrusting (Hull and Dominguez, 2007). However, it is difficult to make a rational bridge between these behavioral observations and the present data. The selective MC4R agonist, THIQ, potentiated the erectile responses induced by ES of the cavernous nerve in mice anesthetized with urethane (Van der Ploeg et al., 2002). As a first pharmacological application of the model described herein, we attempted to assess the proerectile
290
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
facilitator activity of the non-selective melanocortinergic agonist MT-II. MT-II is a highly potent agonist at the MC4R, and the dose used in the present study was shown to potentiate the erectile response induced by ES of the cavernous nerve in rats anesthetized with urethane (Giuliano et al., 2006). Somewhat surprisingly, MT-II (1 mg/kg, i.v.) did not potentiate the reflex erectile responses induced by ES of the DPN in the present experiment. From a penile erection point of view, ES of the cavernous nerve and the DPN can be regarded as similar, the end result of ES of the DPN being a proportionate activation of the postganglionic parasympathetic proerectile neurons which project their axons through the cavernous nerve, at least in rats (Rampin et al., 1994). Indeed, we believed that ES of the DPN was more likely to evidence the proerectile activity of MC4R agonists than ES of the cavernous nerve, since DPN stimulation recruits additional intermediate neurons in the spinal cord and the major pelvic ganglion, all of which are potential targets for MC4R agonists (Van der Ploeg et al., 2002; Allard et al., in press). Therefore, the physiological differences between the generation of erectile responses by ES of the DPN or the cavernous nerve do not give a rational explanation for the lack of effect of MT-II in the present experiment. An alternative potential explanation might be differences between the use of isofluraneanesthesia in mice described herein, compared with urethane in rats (Guliano et al., 2006). Further experiments would be necessary to investigate this hypothesis.
REFERENCES Allard J, Reynolds D, Edmunds NJ (2008) Potentiation of reflex erectile responses in the anaesthetised rat by the selective melanocortin receptor 4 agonist MB243. BJU Int, in press. Allard J, Truitt WA, McKenna KE, Coolen LM (2005) Spinal cord control of ejaculation. World J Urol 23:119 –126. Behr-Roussel D, Darblade B, Oudot A, Compagnie S, Bernabe J, Alexandre L, Giuliano F (2006) Erectile dysfunction in hypercholesterolemic atherosclerotic apolipoprotein E knockout mice. J Sex Med 3:596 – 603. Burnett AL, Johns DG, Kriegsfeld LJ, Klein SL, Calvin DC, Demas GE, Schramm LP, Tonegawa S, Nelson RJ, Snyder SH, Poss KD (1998) Ejaculatory abnormalities in mice with targeted disruption of the gene for heme oxygenase-2. Nat Med 4:84 – 87. Carro-Juarez M, Cruz SL, Rodriguez-Manzo G (2003) Evidence for the involvement of a spinal pattern generator in the control of the genital motor pattern of ejaculation. Brain Res 975:222–228.
Carro-Juarez M, Rodriguez-Manzo G (2000) Sensory and motor aspects of the coital reflex in the spinal male rat. Behav Brain Res 108:97–103. Carro-Juarez M, Rodriguez-Manzo G (2005) Role of genital sensory information in the control of the functioning of the spinal generator for ejaculation. Int J Impot Res 17:114 –120. Champion HC, Bivalacqua TJ, Takimoto E, Kass DA, Burnett AL (2005) Phosphodiesterase-5A dysregulation in penile erectile tissue is a mechanism of priapism. Proc Natl Acad Sci U S A 102:1661–1666. Giuliano F, Bernabe J, Alexandre L, Niewoehner U, Haning H, Bischoff E (2003) Pro-erectile effect of vardenafil: in vitro experiments in rabbits and in vivo comparison with sildenafil in rats. Eur Urol 44:731–736. Giuliano F, Clement P, Droupy S, Alexandre L, Bernabe J (2006) Melanotan-II: Investigation of the inducer and facilitator effects on penile erection in anaesthetized rat. Neuroscience 138:293–301. Hirschfeld RM (2003) Long-term side effects of SSRIs: sexual dysfunction and weight gain. J Clin Psychiatry 64 (Suppl 18):20 –24. Hull EM, Dominguez JM (2007) Sexual behavior in male rodents. Horm Behav 52:45–55. Marson L, McKenna KE (1990) The identification of a brainstem site controlling spinal sexual reflexes in male rats. Brain Res 515:303–308. McKenna KE (2000) Some proposals regarding the organization of the central nervous system control of penile erection. Neurosci Biobehav Rev 24:535–540. McKenna KE, Chung SK, McVary KT (1991) A model for the study of sexual function in anesthetized male and female rats. Am J Physiol 261:R1276 –R1285. Mizusawa H, Hedlund P, Hakansson A, Alm P, Andersson KE (2001) Morphological and functional in vitro and in vivo characterization of the mouse corpus cavernosum. Br J Pharmacol 132:1333–1341. Pacheco P, Camacho MA, Garcia LI, Hernandez ME, Carrillo P, Manzo J (1997) Electrophysiological evidence for the nomenclature of the pudendal nerve and sacral plexus in the male rat. Brain Res 763:202–208. Pescatori ES, Calabro A, Artibani W, Pagano F, Triban C, Italiano G (1993) Electrical stimulation of the dorsal nerve of the penis evokes reflex tonic erections of the penile body and reflex ejaculatory responses in the spinal rat. J Urol 149:627– 632. Rampin O, Giuliano F, Dompeyre P, Rousseau JP (1994) Physiological evidence of neural pathways involved in reflexogenic penile erection in the rat. Neurosci Lett 180:138 –142. Van der Ploeg LH, Martin WJ, Howard AD, Nargund RP, Austin CP, Guan X, Drisko J, Cashen D, Sebhat I, Patchett AA, Figueroa DJ, DiLella AG, Connolly BM, Weinberg DH, Tan CP, Palyha OC, Pong SS, MacNeil T, Rosenblum C, Vongs A, Tang R, Yu H, Sailer AW, Fong TM, Huang C, Tota MR, Chang RS, Stearns R, Tamvakopoulos C, Christ G, Drazen DL, Spar BD, Nelson RJ, MacIntyre DE (2002) A role for the melanocortin 4 receptor in sexual function. Proc Natl Acad Sci U S A 99:11381–11386. Waldinger MD (2007) Premature ejaculation: definition and drug treatment. Drugs 67:547–568.
(Accepted 20 May 2008) (Available online 2 July 2008)
REFLEX PENILE ERECTION IN ANESTHETIZED MICE: AN EXPLORATORY STUDY J. ALLARD* AND N. J. EDMUNDS
eral input from the pelvis (see McKenna, 2000 and Allard et al., 2005, for review). Thus, from a physiological and mechanistic endpoint, penile erection and ejaculation can be viewed as reflex responses modulated by the brain. The afferent limb of this reflex loop consists of sensory inputs from the penis, which are subsequently conveyed in their distal segment by the dorsal penile nerve (DPN) followed by the sensory branch of the pudendal nerve. The efferent limb is more complex, comprising parasympathetic, sympathetic and somatic projections, which are involved in penile erection, the emission and the expulsion phase of ejaculation respectively. Proerectile parasympathetic fibers reach the penis through the cavernous nerve. Sympathetic fibers, that contribute to the emission phase, join the sexual accessory glands through the hypogastric nerve. Finally, the rhythmic contractions of the bulbospongiosus (BS) muscles which are responsible for the expulsion phase of ejaculation, are controlled by motoneurons that project through the motor branch of the pudendal nerve. Most of our knowledge of the pharmacology and neurophysiology of the male sexual reflex has been generated in experiments using male rats. In the urethane-anesthetized rat, McKenna et al. (1991) showed that peripheral stimulation of genital origin, through mechanical stimulation of the penile glans and urethra, could result in penile erection and rhythmic contractions of the BS muscle (see also Carro-Juarez and Rodriguez-Manzo, 2000, 2005; Carro-Juarez et al., 2003). These contractions of the BS muscle are considered a surrogate marker of the expulsion phase of ejaculation. A variation of this original model replaced the mechanical stimuli of the genital tract by direct electrical stimulation (ES) of the penile sensory afferent nerves running in the DPN (Pescatori et al., 1993; Allard et al., in press). A key feature of these experimental models is that complete spinal transection at the thoracic level is necessary for penile erection and the expulsion reflex to be elicited. The likely reason for this is that in intact rats, the spinal network controlling sexual reflexes is tonically inhibited by descending input from the brainstem (Marson and McKenna, 1990). The requirement for spinal transection in this model gives a unique opportunity to specifically interrogate the spinal circuitry controlling sexual reflex. However, from a preclinical research perspective spinal transection precludes the ability to investigate mechanisms that act via central modulation of descending inhibitory input. For example chronic treatment with selective 5-HT reuptake inhibitor modulates descending input causing an increase in ejaculatory latencies in patients suffering from depression or premature ejaculation (see
Pfizer Global Research and Development, Genitourinary TA, Ramsgate Road, Sandwich, Kent, CT13 9NJ, UK
Abstract—Ejaculatory-like rhythmic contractions of the bulbospongiosus (BS) muscle and penile erection can be elicited in the urethane-anesthetized rat, following spinal cord transection, upon electrical stimulation (ES) of the dorsal penile nerve (DPN). The aim of this work was to investigate this reflex in anesthetized mice. Adult C57BL6 mice were anesthetized with isoflurane. The BS muscle and corpus cavernosum were instrumented to allow quantification of the BS muscle electromyographic activity (BS EMG) and intracavernosal pressure respectively. The femoral artery and jugular vein were catheterized to allow measurement of blood pressure and compound administration. ES of the DPN, via bipolar silver electrodes, reliably evoked erectile responses in mice with intact spinal cords. The overall amplitude of the erectile response was frequencyand pulse duration– dependent. Erectile responses were abolished by bilateral cut of the sensory branch of the pudendal nerve. Transection of the spinal cord potentiated the erectile responses and increased the area under the curve of the BS EMG when compared with those animals with intact spinal cords. However, no coordinated rhythmic contractions of the BS muscle during or after the ES could be observed, with or without spinal transection. Melanotan-II failed to enhance the erectile response induced by ES of the DPN, in mice with intact spinal cords. ES of the DPN in isoflurane-anesthetized mice could be a useful model in which to study the interplay between brain and spinal cord in the control of reflex penile erection, and could take advantage of knockout mice models. However, the lack of efficacy of Melanotan-II suggests that further experiments are necessary to confirm the future utility of this model. In contrast to rats, the expulsion reflex could not be reliably elicited in mice with or without spinal transection. This latter finding suggests the existence of fundamental differences in the organization of the spinal network controlling sexual reflexes between rats and mice. © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: dorsal penile nerve, bulbospongiosus muscle, ejaculation, Melanotan-II.
Penile erection and ejaculation are controlled by spinal autonomic centers, the activity of which is dependent upon descending input from supraspinal structures and periph*Corresponding author. Tel: ⫹44-1304646005; fax: ⫹44-1304651987. E-mail address: [email protected] (J. Allard). Abbreviations: AUC, area under the curve; BP, blood pressure; BS, bulbospongiosus; DPN, dorsal penile nerve; EMG, electromyographic; ES, electrical stimulation; ICP, intracavernosal pressure; ICPmax, maximal intracavernosal pressure; ICPmean, mean intracavernosal pressure; MT-II, Melanotan-II.
0306-4522/08$32.00⫹0.00 © 2008 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2008.05.027
283
284
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
Table 1. Overview of the ICPmax/BP, ICPmean/BP and ICPAUC/BP of the erectile responses generated with similar or identical ES consisting of square pulses (6 V, 1 ms) at 3 Hz for 30 s Group #
N
n
ICPmax/BP
ICPmean/BP
ICPAUC/BP (s)
BP (mmHg)
1 2 4 5 6
3/4 4/4 3/4 3/4 3/4
20 12 2 2 3
0.70⫾0.02 0.70⫾0.05 0.77⫾0.01 0.66⫾0.06 0.67⫾0.03
0.42⫾0.01 0.55⫾0.04 0.49⫾0.06 0.41⫾0.12 0.43⫾0.03
14.2⫾0.4 35.4⫾7.7 22.9⫾0.6 16.0⫾4.7 21.3⫾3.0
83.8⫾5.7 65.1⫾2.3 83.0⫾2.5 79.0⫾2.6 75.7⫾3.8
(SCintact) (SCtransected) (Pulse duration) (Frequency)a (MT-II test)
a
The erectile responses mentioned in the frequency group were generated with ES at 4 Hz. Group 3 (which used three mice with intact and cut spinal cord) was omitted from the comparison because the duration of the train of stimulation was consistently shorter (20 s). “N ” corresponds to the number of mice included for quantification/the number of mice originally included in the group. “n” corresponds to the number of responses analyzed per mice.
(Hirschfeld, 2003; Waldinger, 2007) for review). Since reflex contractions of the BS muscle have been reported following urethral distension in urethane-anesthetized mice with intact neuraxis (Burnett et al., 1998), we hypothesized that the sexual reflex, initiated through stimulation of the DPN, could be generated in anesthetized mice with intact spinal cords. The present study aimed to characterize the reflex sexual response of anesthetized mice to ES of the DPN.
EXPERIMENTAL PROCEDURES Surgical preparation C57BL/6 mice (Charles-River, Manston, UK) were anesthetized with isoflurane (4% for induction, 2.5% for surgery, 1.8 –2.0% for recording) with level of anesthesia being assessed by lack of withdrawal reflex upon pinching of the paw. Body temperature was maintained at 37 °C using a homeothermic blanket. The femoral artery was catheterized with a stretched catheter (approximate external diameter, 0.1 mm) for blood pressure (BP) recording. The jugular vein was additionally catheterized for drug delivery when necessary. The skin surrounding the penis was removed and the connective tissue surrounding the right corpus cavernosus carefully removed. A dorsal bundle including the DPNs and dorsal penile vein and arteries was dissected from the underlying corpus spongiosus (it was technically difficult to isolate the DPN alone). A midline incision was made on the scrotal skin and the right BS muscle exposed. A 26 gauge needle connected to a pressure transducer was inserted in the right corpus cavernosus to record intracavernosal pressure (ICP). Both the BP and ICP catheter were filled with heparinized saline (50 IU/ml). Electromyographic (EMG) activity of the BS muscle was measured with bipolar differential recording. The electrode was made from a pair of 26 gauge needles. The tip of the needle was bent to form a hook shape that allowed the needle to stay in place once inserted in the right BS muscle. A bipolar stimulating electrode made of 0.2 mm diameter silver wire was hooked around the dorsal bundle. ESs were performed using a Grass stimulator (Grass S-88, Grass Instruments, Quincy, MA, USA). For spinal transection, an incision was made over the midline on the back at the thoracic level. The membrane connecting the T7 and T8 vertebrae was removed with bone rongeur and the spinal cord transected with spring scissors. Muscle and skin layers were closed separately with staples. For comparison, recordings were also generated with Sprague–Dawley rats (250 –350 g, Charles-River) in almost identical conditions, the difference being that the DPN was bilaterally dissected from the surrounding tissue before being hooked on the stimulating electrodes. In rats or mice, ES of the DPN started 30 min after transection of the spinal cord with the entire protocol lasting a further 30 – 60 min.
Experimental groups Mice were divided into six different experimental groups (Table 1). The reproducibility of successive reflex responses to identical ES (train of square wave pulses of 6 V and 1 ms duration at 3 Hz for 30 s) in mice with spinal cord intact and cut was assessed in group 1 and 2 respectively. The potentiation of reflex responses by spinal transection was assessed in group 3 with ES consisting of trains of square wave pulses of 6 V and 0.1 ms duration at 1, 3 and 9 Hz for 20 s, using mice with spinal cord intact and cut. Preliminary data suggested that such parameters of stimulation would span a range from submaximal to supramaximal erectile responses. A similar experiment was performed in rats using stimulation parameters appropriate for this species. The pulse-duration and frequency dependency of the reflex response were assessed in groups 4 and 5, using train of square wave pulses of 6 V and 0.1–5 ms duration at 3 Hz for 30 s (group 4) or 6 V and 1 or 0.1 ms duration at 0.5– 8 Hz for 30 s (groups 5a and b respectively). ES were performed every 3 min. When different pulsedurations or frequencies were tested, each ES was repeated twice. The potential proerectile activity of Melanotan-II (MT-II, Bachem UK Ltd., Merseyside, UK) on the erectile responses induced by ES of the DPN in mice with spinal cord intact was assessed in group 6. Train of square wave pulses of 6 V and 1 ms duration at 3 Hz for 30 s were performed every 3 min until two consecutive equivalent responses were obtained. Three ES at 1 Hz were then performed (pre-injection ES). MT-II (1 mg/kg) or the corresponding vehicle (saline, 1 ml/kg) was then injected i.v. 30 s after the last ES at 1 Hz. The injection was performed over 30 s. Three ES at 1 Hz were performed again every 3 min, starting 3 min after completion of the i.v. injection. For each mouse, the mean maximal intracavernosal pressure (ICPmax)/BP, mean intracavernosal pressure (ICPmean)/BP and ICPAUC/BP of the three erectile responses after the drug or vehicle injection were expressed as a percentage of the mean before the injection.
Recording analysis The analog signal from the pressure transducer (DTX Plus, Becton Dickinson, Milton Keynes, UK) was amplified using a NL108 amplifier (Digitimer Ltd., Hertfordshire, UK). The EMG signal was amplified and filtered (10 –50 Hz low, 10 –50 kHz high) using NL125 and NL126 module respectively (Digitimer Ltd.). Filters and amplifiers were set into a 900 D Neurolog box (NL905, Digitimer Ltd.). Analog signals were digitized using a DT-P6 converter (Digitimer Ltd.), and fed to a PC. Pressure and EMG signals were sampled at 500 and 5000 Hz respectively. Acquisition and analysis was performed with notochord-hem 4.1.0.38. Erectile responses were quantified using the ratio of the maximal and mean value of the ICP during the response over the corresponding mean BP during the response (ICPmax/BP and ICPmean/BP respectively), and the ratio of the area under the curve (AUC) of the ICP during the response over the correspond-
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
285
Fig. 1. Upper and lower traces correspond to BP and ICP recordings, respectively, in a mouse with an intact spinal cord. ES (square wave pulses of 6 V, 1 ms at 5 Hz for 15 s, bar in abscissa) of the DPN gave rise to immediate erectile response (arrow) that were abolished (arrowhead) after bilateral transection of the sensory branch of the pudendal nerve (PNsbx). Note the change in ICP baseline activity after PNsbx.
ing mean BP (ICPAUC/BP, expressed in s). Since the magnitude of a given erectile response is dependent on BP, all ICP measures were divided by the corresponding BP in order to normalize ICP responses to systemic BP. The beginning and the end of each erectile event were determined manually. ICPmean, ICPmax and AUC were computed from a baseline determined by the mean value of the ICP during the 5 s preceding the cursor delimiting the beginning of the erectile response. The AUC of the rectified BS EMG was computed from the end of the ES (to avoid integrating the stimulation artifact) until the return of BS muscle activity to baseline. For a given parameter, the ratio of the standard deviation over the mean (coefficient of variation) was computed when necessary to evaluate the reproducibility of the reflex response. Data are expressed as mean⫾standard error of the mean.
Statistical analysis Data per group correspond to the mean of each individual mean per parameter. Experimental groups 1, 2, 4 and 5 were generated successively out of an experimental plan and no statistical comparison was performed between and within these groups. In contrast, experiments corresponding to groups 3 and 6 were designed to allow statistical analysis. For these groups, comparison of the
relevant parameters was performed using Student’s t-test. Differences were considered significant when P⬍0.05. All experiments were conducted in accordance with the European Community Council Directive (86/609/EEC), with the UK Animals (Scientific Procedures) Act 1986 and subject to local ethical review. All efforts were made to minimize the number of animals used and their suffering.
RESULTS Exploratory work showed that ES of the DPN resulted in erectile responses in intact mice anesthetized with isoflurane (data not shown). To assess whether the erectile responses induced by ES of the DPN resulted from a spinally mediated reflex response, ES (square wave pulses of 6 V, 1 ms at 5 Hz for 15 s) of the DPN were performed before and after bilateral transection of the sensory branch of the pudendal nerve within the ischiorectal fossa in two mice. Transection of the sensory branch of the pudendal nerve completely suppressed the erectile response to ES of the DPN while increasing baseline ICP
Fig. 2. Upper and lower traces correspond to BP and ICP recordings, respectively. Recordings in A correspond to a mouse with an intact spinal cord, and show the erectile responses to 20 successive ES of the DPN. Erectile responses progressively reach a plateau level and are stable thereafter. Recordings in B correspond to a mouse with the spinal cord transected at the thoracic level, and show the erectile responses to 12 successive ES of the DPN. Note the considerable variation of the baseline ICP and the prolonged duration of the erectile response.
286
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
activity (see Fig. 1 for example of recording). Further exploratory work showed that ES, consisting of a train of square wave pulses of 6 V and 1 ms duration at 3 Hz for 30 s, resulted in maximal or close to maximal erectile responses in three mice (data not shown). These stimulation parameters, which were derived from work in rats in a similar model, were therefore used as a starting point to study the erectile responses to ES of the DPN in isoflurane-anesthetized mice. Groups 1 and 2: reproducibility of successive erectile responses The reproducibility and stability of the erectile responses induced by ES of the DPN were studied in a group of four mice (group 1). One mouse failed to respond to ES of the DPN and was excluded from the quantification. In the remaining three mice, at least 20 consecutive erectile responses could be elicited (Fig. 2A). Analysis of the recordings showed that four to six consecutive ES were necessary for the erectile responses to stabilize (Fig. 2A and Fig. 3). Erectile responses remained relatively stable thereafter, although a slight progressive decrease could be observed after approximately 15 stimulations (Fig. 3). When considering all erectile responses for each individual mouse, the coefficient of variation for the ICPmax/BP never exceeded 14%, making it the most stable parameter characterizing the erectile responses (Fig. 3A, Table 1). ICPmean/BP and ICPAUC/BP were more variable, with coefficient of variation reaching 26% and 28% respectively (Fig. 3B and C, Table 1). The AUC of the rectified BS EMG response (mean value in the group, 22.5⫾3.9 uV·s) was quite variable between stimulations within each mouse, with a coefficient of variation ranging from 29% to 66%. No rhythmic, coordinated contractions of the BS muscle, similar to an ejaculatory pattern, could be detected during or at completion of the ES. Because sexual reflexes are tightly controlled by descending projections from the brain to the lumbo-sacral spinal cord, the reproducibility and stability of the reflex responses induced by ES of the DPN were studied in four mice with complete spinal cord transection at the thoracic level (group 2). Note that since groups 1 and 2 were experimentally and temporally independent of one another, no statistical comparison was attempted between these groups. BP was consistently lower in mice with spinal cord cut compared with spinal cord intact (Table 1). The baseline ICP was noticeably variable at the beginning of the recording in mice with complete spinal cord transection (Fig. 2B). The fluctuation of the baseline tended to decrease throughout the recording, but was still present in three of four of the mice studied at completion of the recording. Following transection at least 12 consecutive erectile responses could be elicited (although more ES could have been performed). For each mouse, the coefficient variation for the ICPmax/BP and ICPmean/BP never exceeded 9% and 17% respectively (Fig. 3A–B, Table 1). In contrast, the ICPAUC/BP was more variable, with a coefficient of variation reaching 38% in two mice (Fig. 3C, Table 1). The AUC of the rectified BS EMG was highly
Fig. 3. Successive erectile responses generated with ES at 3 Hz (pulses at 6 V, 1 ms) for 30 s were quantified in mice with spinal cord intact (filled symbol, 20 responses, three mice) or cut (open symbol, 12 responses, three mice). The maximal and mean value of the erectile response, and the AUC, corrected by the BP, are displayed in A (ICPmax/BP), B (ICPmean/BP) and C (ICPAUC/BP) respectively. The AUC of the BS EMG is displayed in D. The abscissa corresponds to the successive erectile responses. Mice with spinal cord cut displayed larger ICPAUC/BP, and more variable EMG responses of the BS muscle.
variable both between mice (mean value in the group, 98.3⫾73.4 uV·s) and within mice (coefficient of variation ranging from 25% to 89%). No coordinated rhythmic contractions similar to that occurring during an ejaculatory response could be detected during or at completion of the ES in mice (see Fig. 4 for example of BS EMG recording in mice and rat). Group 3: impact of spinal transection on erectile responses Comparison of the erectile responses generated with ES at 3 Hz showed that the ICPmax/BP and ICPmean/BP were similar in mice with spinal cords either transected or intact (Fig. 3A and B, Table 1). In contrast, the ICPAUC/BP of the erectile response was larger in mice with transected spinal cord compared with intact spinal cord (35.4⫾7.7 s and 14.2⫾0.4 s respectively), because of a longer duration of the erectile response. This suggests a suppressive spinal influence on the erectile response to ES of the DNP, which was also revealed by the shorter latency (time between the start of the ES and the rise of the ICP) in mice with spinal cord cut (1.7⫾0.1 s) compared with spinal cord intact (5.7⫾0.4 s).
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
287
Fig. 4. The ICP and BS EMG recordings in response to ES of the DPN in a mouse with spinal cord cut are shown in A and B respectively. The recording in C is a magnification of the recording in B (magnified zone is delimited by the square in B). Note that the ICP rises within seconds after the onset of the ES (A). Magnification of the EMG recording in mice (C) does not show any sign of coordinated activity of the BS muscle. In contrast, magnification of the corresponding BS EMG response in rat with spinal cord cut (D) in identical experimental condition evidences reflex rhythmic coordinated (arrows) contractions of the BS muscle, similar to what is observed during the expulsion phase of ejaculation. In both animals, the response was triggered by ES at 3 Hz.
To further characterize the potentiating effect of spinal transection on the reflex erectile response, a head to head frequency-response comparison was made between intact and spinal cord transected mice (group 3, three mice in each condition). Electrical parameters generating submaximal erectile responses (train of square wave pulses of 6 V, 0.1 ms at 1, 3 and 9 Hz for 20 s) were used to unmask any potentiating effect of spinal transection. Erectile responses generated with ES at 1 Hz appeared dramatically increased in animals with spinal transection (Fig. 5A–C), although this did not prove statistically significant (ICPmax/BP, P⫽0.057; ICPmean/BP, P⫽0.059; ICPAUC/BP, P⫽0.063). However, for erectile responses generated at 9 Hz, the ICPmean/BP and ICPAUC/BP were significantly greater in mice with spinal cord cut than intact (P⫽0.022 and 0.013 respectively). As comparison, a similar experiment was performed in rats. Under isoflurane-anesthesia, the reflex erectile response-frequency curve is rightward shifted by almost one order of magnitude in rat with intact spinal cord compared with mice (data not shown). Therefore, reflex erectile responses were generated in rats with spinal cord intact and transected (four in each group) using train of square wave pulses of 6 V, 1 ms at 1, 3 and 30 Hz for an overall duration of 20 s to generate subthreshold to suprathreshold stimuli (Fig. 5D–F). Erectile responses generated at 1 and 3 Hz were of much greater amplitude in rats with spinal cord transected than intact (Fig. 5D–F). All parameters characterizing erectile responses were significantly increased in rat with spinal cord cut compared with intact with ES at 3 Hz (ICPmax/BP, P⬍0.001; ICPmean/
BP, P⫽0.020; ICPAUC/BP, P⫽0.018). Erectile responses generated with ES at higher frequencies (30 Hz) were comparable in rats with either intact or transected spinal cords. In addition, reflex rhythmic contractions of the BS muscle were observed in four of four rats with spinal cord transection upon ES at 3 and 30 Hz (see Fig. 4D for example of recording). Only one rat with spinal cord intact displayed such contractions upon ES at 30 Hz. Groups 4 and 5: pulse-duration and frequency dependency of the erectile response To further characterize the erectile response to ES of the DPN, the relationship between the duration of the pulse and the resulting erectile response was studied in a group of four mice (group 4). One mouse did not respond to ES of the DPN and was excluded from the quantification. The amplitude of the response increased with the pulse duration from 0.01– 0.5 ms and reached a plateau thereafter (see Fig. 6A for example of recording, and Fig. 6B and C for quantification). Then, the relationship between the frequency of the pulses and the resulting erectile response was studied with train of square wave pulses of 0.1 or 1 ms duration (corresponding to submaximal and maximal pulse duration respectively) in two independent groups of three and four mice respectively (groups 5a and b). One mouse in the 1 ms group failed to respond. All other mice gave consistent responses and were included in the quantification. For both pulse durations, there was a frequency-dependent increase in the amplitude of the erectile response
288
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
the erectile response (see Fig. 6D for example of recording with 0.1 ms duration pulse). Responses induced with 1 ms pulse were greater than responses induced with 0.1 ms pulse at 0.5 and 1 Hz (Fig. 6E and F). Beyond 2 Hz, there was no difference in the responses generated with pulses of 0.1 or 1 ms. The greatest difference was observed at 0.5 Hz, for which the ICPmax/BP was 0.04⫾0.01 and 0.2⫾0.1, and the corresponding ICPAUC/BP 0.7⫾0.2 and 3.1⫾1.2 s with 0.1 and 1 ms pulse respectively. Group 6: effect of MT-II on erectile responses
Fig. 5. ES of the DNP were performed in mice (A–C) or rats (D–F) with spinal cord intact (open bars) or transected (filled bars). ES consisted in square pulses of 6 V, 1 ms at 1, 3 and 30 Hz in rats, and square wave pulses of 6 V, 0.1 ms at 1, 3 and 9 Hz in mice (total duration 20 s for both). Low frequency of ES (1 Hz in mice, 1 and 3 Hz in rat) clearly unraveled the potentiating effect of spinal cord transection of the erectile response in mice and rat. ES at supramaximal frequency (30 Hz in rat, 9 Hz in mice) overcame the effect of spinal transection in rats, but not in mice. Statistical comparison (t-test) was made in each species between spinal cord cut and intact for each frequency (* P⬍0.05; ** P⬍0.01).
from 0.5–2 Hz (see Table 1 for results at 4 Hz). Increasing the frequency above 2 Hz did not increase the amplitude of
Table 1 summarizes the values of the parameters characterizing the erectile responses to ES at 6 V, 1 ms and 3 Hz in the different independent groups generated throughout the present study. Variability of the ICPAUC/BP response between the different groups of animals suggested that the effect of drug treatment would be better observed within individual mice than between different groups. Therefore, erectile responses were generated at 1 Hz before and after the i.v. injection of 1 mg/kg MT-II or the corresponding vehicle (saline) in two groups of five mice. Two mice in the vehicle group failed to respond to ES of the DPN and were excluded. The ICPmax/BP, ICPmean/BP and ICPAUC/BP of the erectile responses after the injection of MT-II (n⫽5) represented 93⫾5, 94⫾16 and 103⫾17% of the response before the injection, compared with 97⫾4, 93⫾8 and 104⫾12% after the corresponding vehicle injection (n⫽3, see Table 1 for basal responses). Thus, there was no proerectile facilitator effect of MT-II 1 mg/kg in the present model.
Fig. 6. Recording in A (BP and ICP corresponding to upper and lower trace respectively) illustrates the effect of increasing pulse duration (from 0.01–5 ms) on the erectile responses to ES (pulses of 6 V delivered at 3 Hz for 30 s) of the DPN in mice with spinal cord intact. Quantification of similar recordings in three mice shows the pulse-duration dependency of the ICPmax/BP and ICPmean/BP (square and triangle respectively, B) and the ICPAUC/BP (C) of the erectile response. Recording in D illustrates the effect of increasing frequency (from 0.5– 8 Hz) on the erectile responses to ES (pulses of 6 V, 1 ms for 30 s) of the DPN. Similar recordings were generated in two groups of three mice using pulses of 0.1 ms or 1 ms duration. Black and gray traces (1 and 0.1 ms respectively) illustrate the corresponding frequency dependency of the ICPmax/BP and ICPmean/BP (square and triangle respectively, E) and the ICPAUC/BP (F) of the erectile response.
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
DISCUSSION The exploratory work presented herein demonstrates that ES of the DPN induces frequency dependent reflex erectile responses in isoflurane-anesthetized C57BL/6 mice with intact spinal cords. These reflex erectile responses could be potentiated by complete transection of the spinal cord. One consideration regarding the present experiments concerns the lack of erectile response to ES of the DPN in 20% of the mice. We have no rational explanation for this finding at present, since there were no apparent differences in surgical preparation or depth of anesthesia between groups. Attempts to use other anesthetics (urethane 2 g/kg, ketamine/medetomidine 75 mg/kg/1 mg/kg and ␣-chloralose/ketamine 120 mg/kg/100 mg/kg delivered i.p., and propofol 13 mg/kg/min for 2 min followed by 2–3 mg/ kg/min delivered i.v.) were unsuccessful (data not shown). In our hand it was extremely difficult to obtain a stable plan of anesthesia with these anesthetics, and they did not reduce the number of non-responding mice. Thus, in our opinion, isoflurane is the most practical and reliable anesthetic to induce erectile response to ES of the DPN. There is no functional description of the sacral plexus and pudendal nerve anatomy in mice. The sensory branch of the pudendal nerve, which is the proximal end of the DPN, was therefore identified according to the description made in rat by Pacheco and co-workers (1997). Bilateral transection of the identified sensory branch of the pudendal nerve eliminated the erectile response to ES of the DPN. This suggests that the erectile responses generated in the present experiment involved a reflex loop comprising the DPN/sensory branch of the pudendal nerve (afferent arm) and the pelvic nerve/cavernous nerve (efferent arm). The increased ICP baseline activity that was observed after cutting the nerve might be due to transection-induced firing within the proximal end that would partially activate the reflex loop (R. D. Johnson, personal communication). Further experiments involving transection of the efferent arm (i.e. the pelvic and/or cavernous nerve) would be necessary to confirm the existence of this reflex erectile loop in mice, as has been previously established in rats (Rampin et al., 1994). The erectile responses to ES of the DPN were pulse duration-dependent for a set frequency. The pulse-duration dependency might correspond to a progressive increase in the number of fibers recruited by the ES and/or a change in the subtype of fibers (from A to A⫹C) recruited by the ES within the DPN. Concomitant recording of conduction velocity (neurogram) and erectile response for a given parameter of ES would be necessary to determine whether A fibers alone, or A and C fibers are involved in the generation of reflex erectile response in this model. The erectile responses were also frequency-dependent for a set square wave pulse voltage/duration. The presence of a graded response to increasing frequency of stimulation suggests that this model is suitable to assess ability of a drug to facilitate erections (see later), in the same way as ES of the cavernous nerve has been used to illustrate the proerectile activity of PDE5 inhibitors in mice or rats (see
289
for example Mizusawa et al., 2001; Giuliano et al., 2003; Champion et al., 2005). Interestingly, the frequency necessary to reach the plateau response upon ES of the DPN (2–3 Hz) was consistently lower than the frequency necessary to reach the plateau response upon ES of the cavernous nerve in similar conditions of anesthesia in the same mice strain (10 –15 Hz) (Behr-Roussel et al., 2006). This suggests a consistent amplification of the afferent signal, probably at the level of the spinal cord. Burnett et al. (1998) claimed that urethral distention induced reflex ejaculatory-like rhythmic contractions of the BS muscle (expulsion reflex) in urethane-anesthetized C57BL/6x129/SvEv mice with the spinal cord intact. In the present study, no expulsion reflex could be observed following ES of the DPN in isoflurane-anesthetized C57BL/6 mice with spinal cord intact. We believe that the present results are indeed in agreement with the report of Burnett and co-workers, as close examination of their data (see Fig. 2a) does not evidence the coordinated contractions of the BS muscle normally associated with ejaculation, in response to urethral stimulation. Nevertheless, an apparently greater non-coordinated activity of the BS muscle was observed following the stimulation in the study of Burnett and coworkers, which was not observed herein. This difference might be explained by the difference in strain or anesthetic used, or the type of stimulation. In agreement with the general view that descending projections from the brainstem exert a tonic inhibition on the spinal structures controlling reflex erection (McKenna, 2000), spinal cord transection potentiated the erectile response to ES of the DPN in mice. Nevertheless, the present study suggests differences in the spinal network controlling reflexive erections between rats and mice. In animals with spinal transection, the latency of the ICP rise in mice was much shorter when compared with rats (within seconds of the start of the ES and at completion of the ES respectively). Furthermore, in contrast to urethane-anesthetized rats (Marson and McKenna, 1990; McKenna et al., 1991; Pescatori et al., 1993), spinal cord transection did not allow the generation of rhythmic contractions of the BS muscle in response to ES of the DPN in isoflurane-anesthetized mice (present study). This difference cannot be solely attributed to the properties of isoflurane, as we have shown here that in isoflurane-anesthetized rats with spinal transection, ES of the DPN results in rhythmic contractions of the BS muscle. The differences observed herein between mice and rat sexual responses could well be a reflection of differences in their sexual behavior. During sexual behavior, rats require several brief intromissions to reach ejaculation, whereas mice require prolonged intromission with repeated thrusting (Hull and Dominguez, 2007). However, it is difficult to make a rational bridge between these behavioral observations and the present data. The selective MC4R agonist, THIQ, potentiated the erectile responses induced by ES of the cavernous nerve in mice anesthetized with urethane (Van der Ploeg et al., 2002). As a first pharmacological application of the model described herein, we attempted to assess the proerectile
290
J. Allard and N. J. Edmunds / Neuroscience 155 (2008) 283–290
facilitator activity of the non-selective melanocortinergic agonist MT-II. MT-II is a highly potent agonist at the MC4R, and the dose used in the present study was shown to potentiate the erectile response induced by ES of the cavernous nerve in rats anesthetized with urethane (Giuliano et al., 2006). Somewhat surprisingly, MT-II (1 mg/kg, i.v.) did not potentiate the reflex erectile responses induced by ES of the DPN in the present experiment. From a penile erection point of view, ES of the cavernous nerve and the DPN can be regarded as similar, the end result of ES of the DPN being a proportionate activation of the postganglionic parasympathetic proerectile neurons which project their axons through the cavernous nerve, at least in rats (Rampin et al., 1994). Indeed, we believed that ES of the DPN was more likely to evidence the proerectile activity of MC4R agonists than ES of the cavernous nerve, since DPN stimulation recruits additional intermediate neurons in the spinal cord and the major pelvic ganglion, all of which are potential targets for MC4R agonists (Van der Ploeg et al., 2002; Allard et al., in press). Therefore, the physiological differences between the generation of erectile responses by ES of the DPN or the cavernous nerve do not give a rational explanation for the lack of effect of MT-II in the present experiment. An alternative potential explanation might be differences between the use of isofluraneanesthesia in mice described herein, compared with urethane in rats (Guliano et al., 2006). Further experiments would be necessary to investigate this hypothesis.
REFERENCES Allard J, Reynolds D, Edmunds NJ (2008) Potentiation of reflex erectile responses in the anaesthetised rat by the selective melanocortin receptor 4 agonist MB243. BJU Int, in press. Allard J, Truitt WA, McKenna KE, Coolen LM (2005) Spinal cord control of ejaculation. World J Urol 23:119 –126. Behr-Roussel D, Darblade B, Oudot A, Compagnie S, Bernabe J, Alexandre L, Giuliano F (2006) Erectile dysfunction in hypercholesterolemic atherosclerotic apolipoprotein E knockout mice. J Sex Med 3:596 – 603. Burnett AL, Johns DG, Kriegsfeld LJ, Klein SL, Calvin DC, Demas GE, Schramm LP, Tonegawa S, Nelson RJ, Snyder SH, Poss KD (1998) Ejaculatory abnormalities in mice with targeted disruption of the gene for heme oxygenase-2. Nat Med 4:84 – 87. Carro-Juarez M, Cruz SL, Rodriguez-Manzo G (2003) Evidence for the involvement of a spinal pattern generator in the control of the genital motor pattern of ejaculation. Brain Res 975:222–228.
Carro-Juarez M, Rodriguez-Manzo G (2000) Sensory and motor aspects of the coital reflex in the spinal male rat. Behav Brain Res 108:97–103. Carro-Juarez M, Rodriguez-Manzo G (2005) Role of genital sensory information in the control of the functioning of the spinal generator for ejaculation. Int J Impot Res 17:114 –120. Champion HC, Bivalacqua TJ, Takimoto E, Kass DA, Burnett AL (2005) Phosphodiesterase-5A dysregulation in penile erectile tissue is a mechanism of priapism. Proc Natl Acad Sci U S A 102:1661–1666. Giuliano F, Bernabe J, Alexandre L, Niewoehner U, Haning H, Bischoff E (2003) Pro-erectile effect of vardenafil: in vitro experiments in rabbits and in vivo comparison with sildenafil in rats. Eur Urol 44:731–736. Giuliano F, Clement P, Droupy S, Alexandre L, Bernabe J (2006) Melanotan-II: Investigation of the inducer and facilitator effects on penile erection in anaesthetized rat. Neuroscience 138:293–301. Hirschfeld RM (2003) Long-term side effects of SSRIs: sexual dysfunction and weight gain. J Clin Psychiatry 64 (Suppl 18):20 –24. Hull EM, Dominguez JM (2007) Sexual behavior in male rodents. Horm Behav 52:45–55. Marson L, McKenna KE (1990) The identification of a brainstem site controlling spinal sexual reflexes in male rats. Brain Res 515:303–308. McKenna KE (2000) Some proposals regarding the organization of the central nervous system control of penile erection. Neurosci Biobehav Rev 24:535–540. McKenna KE, Chung SK, McVary KT (1991) A model for the study of sexual function in anesthetized male and female rats. Am J Physiol 261:R1276 –R1285. Mizusawa H, Hedlund P, Hakansson A, Alm P, Andersson KE (2001) Morphological and functional in vitro and in vivo characterization of the mouse corpus cavernosum. Br J Pharmacol 132:1333–1341. Pacheco P, Camacho MA, Garcia LI, Hernandez ME, Carrillo P, Manzo J (1997) Electrophysiological evidence for the nomenclature of the pudendal nerve and sacral plexus in the male rat. Brain Res 763:202–208. Pescatori ES, Calabro A, Artibani W, Pagano F, Triban C, Italiano G (1993) Electrical stimulation of the dorsal nerve of the penis evokes reflex tonic erections of the penile body and reflex ejaculatory responses in the spinal rat. J Urol 149:627– 632. Rampin O, Giuliano F, Dompeyre P, Rousseau JP (1994) Physiological evidence of neural pathways involved in reflexogenic penile erection in the rat. Neurosci Lett 180:138 –142. Van der Ploeg LH, Martin WJ, Howard AD, Nargund RP, Austin CP, Guan X, Drisko J, Cashen D, Sebhat I, Patchett AA, Figueroa DJ, DiLella AG, Connolly BM, Weinberg DH, Tan CP, Palyha OC, Pong SS, MacNeil T, Rosenblum C, Vongs A, Tang R, Yu H, Sailer AW, Fong TM, Huang C, Tota MR, Chang RS, Stearns R, Tamvakopoulos C, Christ G, Drazen DL, Spar BD, Nelson RJ, MacIntyre DE (2002) A role for the melanocortin 4 receptor in sexual function. Proc Natl Acad Sci U S A 99:11381–11386. Waldinger MD (2007) Premature ejaculation: definition and drug treatment. Drugs 67:547–568.
(Accepted 20 May 2008) (Available online 2 July 2008)