Alpha-Adrenergic Receptor Blockade by Phentolamine Increases the Ef®Cacy of Vasodilators in Penile Corpus Cavernosum

Alpha-adrenergic receptor blockade by phentolamine increases the ef®cacy of vasodilators in penile corpus cavernosum

NN Kim1*, I Goldstein1, RB Moreland1 and AM Traish1

1Department of Urology, Boston University School of Medicine, Boston, Massachusetts 02118, USA

Penile trabecular smooth muscle tone, a major determinant of erectile function, is highly regulated by numerous inter- and intracellular pathways. The interaction between pathways mediating contraction and relaxation has not been studied in detail. To this end, we investigated the functional effects of alpha adrenergic receptor blockade with phentolamine and its interaction with vasodilators (sildena®l, vasoactive intestinal polypeptide (VIP) and PGE1) that elevate cyclic nucleotides on penile cavernosal smooth muscle contractility. In organ bath preparations of cavernosal tissue strips contracted with phenylephrine, phentolamine signi®cantly enhanced relaxation induced by sildena®l, VIP and PGE1. Sildena®l, VIP or PGE1 also signi®cantly enhanced relaxation induced by phentolamine in cavernosal tissue strips contracted with phenylephrine. To study the effects of alpha adrenergic receptor blockade and modi®cation of cyclic nucleotide metabolism during active neurogenic input, cavernosal tissue strips in organ bath preparations were contracted with the non-adrenergic agonist endothelin-1 and subjected to electrical ®eld stimulation (EFS) in the absence or presence of phentolamine and=or sildena®l. EFS (5±40 Hz) typically caused biphasic relaxation and contraction responses. Phentolamine alone enhanced relaxation and reduced or prevented contraction to EFS. Sildena®l enhanced relaxation to EFS at lower frequencies ( 5 Hz). The combination of phentolamine and sildena®l enhanced EFS-induced relaxation at all frequencies tested. EFS, in the presence of 10 nM phentolamine and 30 nM sildena®l, produced enhanced relaxation responses which were quantitatively similar to those obtained in the presence of 50 nM sildena®l alone. Thus, blockade of a-adrenergic receptors with phentolamine increases the ef®cacy of cyclic nucleotide-dependent vasodilators. Further- more, phentolamine potentiates relaxation and attenuates contraction in response to endogenous neurotransmitters which are released during EFS. These ®ndings suggest that antagonism of a- adrenergic signaling enables other independent relaxatory pathways to predominate within penile trabecular smooth muscle. International Journal of Impotence Research (2000) 12, Suppl 1, S26±S36

Keywords: alpha adrenergic antagonist; sildena®l; VIP; prostaglandin E1

Introduction Alpha-2 adrenergic receptors inhibit adenylate cyclase activity via Gi protein and have also been shown to modulate calcium in¯ux either by G- The signaling pathways stimulated by alpha adren- protein coupled or G-protein independent mechan- ergic receptors and intracellular cyclic nucleotides isms.8±10 These receptors are also implicated in the (cAMP and cGMP) are important regulators of penile activation of phospholipase C, phospholipase D, pot- cavernosal smooth muscle tone. Contraction of assium channels and Na=H exchange.10 These corpus cavernosum trabecular smooth muscle to alpha-adrenergic receptor-mediated events result in catecholamines is mediated by post-junctional elevated intracellular calcium or increased calcium 1±7 alpha-1 and alpha-2 adrenergic receptors. The sensitivity and ultimately enhance actin-myosin binding of catecholamines to alpha-1 adrenergic cross-bridge cycling to increase smooth muscle tone. receptors stimulates interaction with Gq=11 protein, Thus, alpha-adrenergic receptors are crucial in the causing the activation of phospholipase Cb1 and process of detumescence and may be important in 8,9 subsequent production of IP3 and diacylglycerol. maintaining penile ¯accidity.11±14 Multiple receptor systems activate cyclic nucleotide-mediated relaxation. In addition to the direct *Correspondence: Dr NN Kim, Department of Urology, Boston activation of guanylate cyclase by nitric oxide,15 two University School of Medicine 700 Albany St. W607 Boston, MA 02118. agonists (among others) which stimulate cAMP E-mail: [email protected] production in penile trabecular smooth muscle are Received 1 November 1999; accepted 23 November 1999 prostaglandin E1 (PGE1) and vasoactive intestinal Phentolamine and vasodilators NN Kim et al 27 polypeptide (VIP).14 EP receptors mediate the ments, Quincy, MA) and immersed in 25 ml baths cellular response to PGE1 and are categorized into of PSS at 37 C which were aerated with 5% CO2, four pharmacologic subclasses (EP1±EP4), with iso- 19% O2, 76% N2. Tissues were treated with 3 mM form variants in each subclass which may arise from indomethacin and progressively stretched until alternative splicing of a single gene product.16±19 optimal isometric tension was reached, as described 30 VIP, EP2,EP3II and EP4 receptors mediate smooth previously. Tissues were then subjected to one of muscle relaxation by coupling to Gs protein and several experimental protocols, as outlined below. stimulating adenylate cyclase to increase intra- All protocols included 3 mM indomethacin to inhibit cellular cAMP.14; 18 23 In the clinical setting, prostanoid production. systemic administration of alpha-adrenergic ant- agonists facilitates penile erection and has been associated with prolonged erection or priapism.11±14,24±28 In addition, orally administered Dose responses doxazosin, an alpha-adrenergic receptor antagonist, has been reported to increase the ef®cacy of Cavernosal tissue strips in organ bath preparations intracavernosal vasodilator therapy with alprostadil were contracted with 1 mM phenylephrine or 40 nM (PGE1) in men with moderate to severe erectile endothelin-1 as speci®ed in the results section. dysfunction.29 While it may be postulated that the Tissues were incubated with vehicle or 5±50 nM relative predominance of any of the regulatory phentolamine mesylate (Research Biochemicals, mechanisms could favor penile tumescence or Inc., Natick, MA) for 20 min and then exposed to detumescence, the interaction between contractile increasing concentrations of sildena®l citrate (P®zer and relaxatory pathways has not been studied in Ltd., Sandwich, Kent, UK), vasoactive intestinal detail. Thus, the objective of this study was to polypeptide (VIP; Peninsula Laboratories, San Car- investigate the functional effects of alpha-adrenergic los, CA) or prostaglandin E1 (PGE1; Cayman Chemi- receptor blockade and its interaction with vasodila- cal Co., Ann Arbor, MI). In parallel experiments, tors (sildena®l, VIP and PGE1) that elevate cyclic cavernosal tissue strips were contracted with 1 mM nucleotides on penile corpus cavernosum smooth phenylephrine or 40 nM endothelin-1, incubated muscle contractility. with sildena®l (0±20 nM), VIP (0±100 nM) or PGE1 (0±500 nM) and exposed to increasing concentra- Methods tions of phentolamine. Tissue procurement

Electrical ®eld stimulation All protocols for studies involving human and rabbit tissues were approved by the Institutional Review Board for Human Studies and the Institutional Cavernosal tissue strips in organ bath preparations Animal Care and Use Committee at Boston Uni- were contracted with 40 nM endothelin-l and sub- versity School of Medicine. Human penile corpus jected to electrical ®eld stimulation. As described cavernosum tissue was obtained from patients previously,30 EFS was accomplished by means of diagnosed with erectile dysfunction and undergoing two platinum plate electrodes, positioned on either penile prosthesis implantation. Cavernosal biopsies side of the tissue and a current ampli®er in series were transported to the research laboratory in ice- with a square pulse stimulator (Model SD9; Grass cold physiological salt solution (PSS; 118.3 mM Instruments, Quincy, MA). An oscilloscope was NaCl, 4.7 mM KCl, 0.6 mM MgSO4, 1.2 mM used to monitor stimulation parameters. Each KH2PO4, 2.5 mM CaCl2, 25 mM NaHCO3, 0.026 mM stimulation period lasted 20 s with trains of square CaNa2EDTA, 11.1 mM glucose) for immediate study waves having a pulse duration of 0.5 ms, and a and cut into strips (3 6 3 6 10 mm) for use in organ potential difference of 10 V. Frequency was varied bath experiments. Rabbit corpus cavernosum tissue from 0.5 to 40 Hz. Electrical stimulation with these was obtained from euthanized male New Zealand parameters elicits tetrodotoxin sensitive responses White rabbits (3.0±3.5 kg body weight), as pre- in corpus cavernosum, indicating a speci®c nerve- viously described.30 Cavernosal tissue from each mediated action.31 In some experiments, tissue rabbit was cut into four strips for organ bath strips were treated with 1 mM atropine and 10 mM experiments. bretylium to isolate non-adrenergic, non-cholinergic responses. After the initial set of stimulations, Measurement of changes in isometric tension tissues were treated with phentolamine alone, sildena®l alone or a combination of phentolamine and sildena®l at the indicated concentrations. After Human and rabbit cavernosal tissues were mounted stable tone was achieved, electrical stimulations onto force transducers (Model FT03; Grass Instru- were repeated.

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 28 Analysis of data 1 mM phenylephrine, addition of a single concentration of phentolamine caused a sustained relaxation. At all concentrations tested, phentolamine en- At the end of each experiment, all tissue strips were hanced relaxation to sildena®l, although concentra- treated with 10 mM papaverine and 10 mM nitroprus- tions above 10 nM did not cause further side to induce maximal relaxation (100%). For dose improvement (Figure 1A and Table 1). When the responses, the `relaxation index' was determined by data were normalized to account for the initial crude integration of the area over the relaxation relaxation caused by phentolamine treatment prior curve. The relaxation index of a given dose response to sildena®l exposure, this potentiation did not obtained in treated tissue was compared to that persist (Figure 2A). As shown in Table 1, in tissues obtained in untreated (control) tissue by t-test. EC50 contracted with the non-adrenergic vasoconstrictor values were estimated by constructing pseudo-Hill endothelin-1, phentolamine (5±20 nM) caused a plots. Using the equation describing the line of slight relaxation. However, no signi®cant potentia- linear regression, we determined the concentration tion of sildena®l-induced relaxation was observed at of drug (x) at which log[R=(Rmax 7 R)] 0(y); where any concentration of phentolamine tested. As R percent of maximal relaxation response and indicated in Table 1, EC50 values and relaxation Rmax maximal relaxation (i.e. 100%). For electrical indices were not signi®cantly altered with phento- ®eld stimulation studies, the response at each lamine treatment. frequency of stimulation in the presence of drug Since the effects of VIP are readily reversed by was compared to control responses using paired t- extensive washing, the control response was ob- test. Responses were further analyzed by determin- tained ®rst for each tissue strip and the response ing the ratios of the responses in the presence and was then repeated in the presence of phentolamine. absence of drug treatment and comparing these In this set of experiments, VIP caused relaxation in ratios with those of the time control. For all t-tests phenylephrine-contracted cavernosal tissue strips (paired and unpaired), comparisons were deemed to with varying ef®cacy. In the absence of phentol- be statistically signi®cant if the two-tailed P-value amine, relaxation responses in rabbit cavernosal was less than or equal to 0.05 (indicated by an tissue exhibited greater consistency with a mean asterisk). EC50 value of 56 nM. EC50 values for control responses in human cavernosal tissue varied from Results 48.5 to 214.7 nM (Table 2). Subsequent VIP dose responses in the presence of vehicle were not Effect of phentolamine on the ef®cacy of sildena®l, signi®cantly different from the initial responses, indicating that repeated exposure to VIP alone did VIP or PGE1 not alter the sensitivity of the tissue. In both rabbit and human tissues treated with phentolamine, VIP Relaxation dose responses to sildena®l, VIP or PGE1 responses were progressively potentiated, relative to were performed in the absence or presence of the dose of phentolamine which was used (Table 2 phentolamine in rabbit and human penile caver- and Figure 1B). In addition, phentolamine caused a nosal tissue strips. In tissue strips contracted with larger enhancement of VIP-induced relaxation in

Figure 1 Effect of phentolamine on vasodilator-induced relaxation. Organ bath preparations of rabbit and human penile cavernosal tissue strips were contracted with 1 mM phenylephrine and treated with vehicle (control) or phentolamine (PA) at the indicated concentrations. Tissue strips were then exposed to increasing concentrations of sildena®l (A), VIP (B) or prostaglandin E1 (C).

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al

Table 1 Effect of phentolamine on sildena®l-induced relaxation in tissues contracted with phenylephrine (data are mean Æ SE) 29

Treatment N EC50 (nM) Relaxation index Percent change Phenylephrine contracted Control 11 35.2 443.5 Æ 24.9 ± 5 nM phentolamine 12 9.7* 524.0 Æ 35.0 18.2 10 nM phentolamine 12 2.6* 637.0 Æ 17.0* 43.6 20 nM phentolamine 12 2.9* 597.4 Æ 26.0* 34.7 Endothelin contracted Control 11 39.2 434.7 Æ 32.5 ± 5 nM phentolamine 8 26.1 466.7 Æ 43.8 7.4 10 nM phentolamine 12 24.4 472.4 Æ 35.4 8.7 20 nM phentolamine 14 34.9 442.6 Æ 39.9 1.8

*P 0.05, compared to control.

Figure 2 Data in Figure 1 normalized by subtracting the initial relaxation response caused by phentolamine prior to the addition of vasodilators.

Table 2 Effect of phentolamine on VIP-induced relaxation in penile corpus cavernosum contracted with phenylephrine (Data are mean Æ SE)

Treatment N EC50(nM) Relaxation index Percent change Rabbit corpus cavernosum Control 5 58.8 Æ 26.0 307.9 Æ 25.9 Vehicle 5 100.9 Æ 38.2 299.2 Æ 22.9 7 2.8 Control 4 50.3 Æ 20.4 318.7 Æ 37.9 10 nM phentolamine 4 61.0 Æ 29.4 457.3 Æ 34.6 43.5 Control 5 66.4 Æ 26.7 303.4 Æ 30.9 20 nM phentolamine 5 41.3 Æ 20.1 509.6 Æ 46.4 68.0 Control 4 48.3 Æ 18.1 302.6 Æ 29.1 40 nM phentolamine 4 21.9 Æ 10.4 544.5 Æ 30.6 79.9 Human corpus cavernosum Control 5 48.5 Æ 13.5 127.7 Æ 23.6 Vehicle 5 89.1 Æ 33.0 150.5 Æ 15.5 17.8 Control 6 214.7 Æ 64.7 160.6 Æ 26.5 10 nM phentolamine 6 82.4 Æ 34.0* 322.2 Æ 374* 100.6 Control 6 182.3 Æ 38.0 175.5 Æ 20.3 20 nM phentolamine 6 67.0 Æ 13.9* 385.5 Æ 12.7* 119.6 Control 5 214.3 Æ 81.1 156.2 Æ 23.2 40 nM phentolamine 5 80.0 Æ 32.3* 457.8 Æ 43.0* 193.1

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 30 Table 3 Effect of phentolamine on PGE1-induced relaxation in tissues contracted with phenylephrine (data are mean Æ SE)

Treatment N EC50 (nM) Relaxation index Percent change Control 5 169.5 335 Æ 18 ± 10 nM phentolamine 5 46.0* 432 Æ 38 28.9 20 nM phentolamine 5 23.9* 469 Æ 44* 40.0 50 nM phentolamine 5 3.4* 629 Æ 97* 87.7

*P 0.05, compared to control.

Table 4 Effect of sildena®l, VIP and PGE1 on phentolamine-induced relaxation in penile cavernosal tissues contracted with phenylephrine (data are mean Æ SE)

Treatment N EC50 (nM) Relaxation index Percent change

Sildena®l Control 9 11.7 533.7 Æ 47.1 ± 5 nM Sildena®l 9 0.9* 726.5 Æ 38.6* 36.1 10 nM Sildena®l 9 0.6* 755.3 Æ 54.7* 41.5 20nMSildena®l 9 0.8* 761.1 Æ 43.1* 42.6 Vasoactive intestinal polypeptide Control 6 19.6 528.8 Æ 19.5 ± 10 nM VIP 6 10.5 660.2 Æ 13.7* 24.8 50 nM VIP 6 22.7 600.4 Æ 28.2* 13.5 100 nM VIP 6 16.2 637.7 Æ 17.0* 20.6 Prostaglandin E1 Control 2 369.5 202.2 Æ 13.6 ± 50 nM PGE1 2 38.6 340.9 Æ 33.7 68.6 500 nM PGE1 2 24.9 386.2 Æ 19.5 91.0

*P 0.05, compared to control.

Figure 3 Effect of vasodilators on phentolamine-induced relaxation. Organ bath preparations of rabbit (A B) and human (C) penile cavernosal tissue strips were contracted with 1 mM phenylephrine and treated with vehicle (control), sildena®l (SIL), vasoactive intestinal polypeptide (VIP) or prostaglandin E1 (PGE) at the indicated concentrations. Tissue strips were then exposed to increasing concentrations of phentolamine.

human tissue, relative to the rabbit. Unlike sildena- Similar to VIP, PGE1-induced relaxation was ®l, signi®cant differences between VIP-induced signi®cantly enhanced in a dose-dependent manner relaxation in the absence and presence of 40 nM by the addition of 10±50 nM phentolamine (Table 3 phentolamine persisted even when the data were and Figure 1C). Phentolamine, at 50 nM, enhanced normalized to account for the initial relaxation PGE1-mediated relaxation by approximately 50-fold. caused by phentolamine treatment prior to VIP This potentiation also persisted (although to a lesser exposure (Figure 2B). degree) after the data were normalized to account for

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 31 the effect of phentolamine before addition of PGE1 500 nM PGE1 elicited relaxation ranging from 0 to (Figure 2C). Dose responses to PGE1 were performed 60% of the maximal response. Nevertheless, only in human cavernosal tissue because rabbit signi®cant enhancement was observed when phen- penile corpus cavernosum responds poorly to tolamine dose responses were performed in the relaxatory prostanoids. presence of PGE1. Interestingly, this potentiation occurred even though the average relaxation response to PGE1 was minimal (Figure 3C). Thus, the enhancement persisted after normalization of the Effect of sildena®l, VIP and PGE1 on phentolamine- data (Figure 4C). induced relaxation

To further study the interaction of adrenergic re- Electrical ®eld stimulation studies ceptor antagonism and cyclic nucleotide-mediated smooth muscle relaxation, we performed dose responses to phentolamine in the absence or Penile cavemosal tissue was subjected to EFS at presence of sildena®l, VIP or PGE1. These vasodila- varying frequencies to cause the release of endogen- tors were used at concentrations which produced ous neurotransmitters. This protocol enabled us to mild to moderate trabecular smooth muscle relaxa- study the effects of adrenergic receptor blockade and tion. On average, sildena®l (5±20 nM) enhanced modi®cation of cyclic nucleotide metabolism during relaxation to phentolamine by 1.4-fold and de- active neurogenic input without the complications creased the EC50 of phentolamine by a factor of 15 of systemic factors that may arise from in vivo (Table 4 and Figure 3A). Higher concentrations of experimental models. In addition to phentolamine, sildena®l did not signi®cantly increase the potentia- the mechanism of action of sildena®l enabled us to tion observed at 5 nM. When the data were normal- examine the functional interaction between adre- ized to account for the initial relaxation caused by nergic receptor-mediated contraction and cGMP- sildena®l treatment prior to phentolamine exposure mediated relaxation. The exogenous agonists VIP (Figure 4A), the potentiation did not persist. and PGE1 were not studied in this paradigm. All VIP and PGE1 also enhanced relaxation to tissue strips were contracted with endothelin-1 phentolamine (Table 4 and Figure 3). The enhance- (40 nM) to exclude exogenous adrenergic stimula- ment by VIP was somewhat less than the potentiation. In the absence of vasoactive modi®ers, EFS tion by sildena®l and also did not persist when the elicited characteristic biphasic responses consisting data were normalized to account for the initial of a primary relaxation, followed by a contraction. relaxation produced by VIP prior to the addition of Contractile responses became more prevalent and phentolamine (Figure 4B). Responses to PGE1 were relaxatory responses decreased as a function of variable in human corpus cavernosum tissue ob- increasing frequency (Figures 5 and 7). Immediately tained from different patients. Treatment with 50 or after EFS was terminated, tissues sometimes exhib-

Figure 4 Data in Figure 3 normalized by subtracting the initial relaxation response caused by sildena®l (A), VIP (B) or PGE1 (C) prior to the addition of phentolamine.

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 32

Figure 5 Effect of phentolamine on nerve-mediated responses. Rabbit cavernosal tissue strips in organ baths were contracted with 40 nM endothelin-1 (ET-1) and subjected to electrical ®eld stimulation at varying frequencies (0.5±40 Hz). After washout, tissue strips were treated with 20 nM phentolamine for 30 min and the frequency responses were repeated. Papaverine (Pap) was added at the end of the experiment to determine baseline tension. Shown are representative polygraph tracings from two separate tissue strips.

Figure 6 Summarized data for rabbit cavernosal tissue relaxation responses to electrical ®eld stimulation in the absence or presence of phentolamine. Rabbit cavernosal tissue strips in organ baths were contracted with 40 nM endothelin-1 (ET-1) and subjected to electrical ®eld stimulation at the indicated frequencies (white bars). After washout, tissue strips were treated with vehicle (Time control), 10 nM phentolamine or 20 nM phentolamine for 30 min and the frequency responses were repeated (black bars). *P 0.05, relative to ®rst response.

ited a slower, but transient, secondary relaxation. EFS-induced relaxation to a greater degree than the The effect of repeated stimulation with no drug time control, although the effect was more pro- treatment was assessed in the `time control.' Inter- nounced at lower frequencies ( 5 Hz; data not estingly, relaxation to EFS increased signi®cantly shown). The responses to EFS with the combined with the second series of stimulations at frequencies treatment of phentolamine and sildena®l enhanced of 0.5±5 Hz (Figure 6). Higher frequencies ( > 5 Hz) relaxation in a dose dependent manner to a greater were not signi®cantly different. degree than the time control at all frequencies Phentolamine (10±20 nM) potentiated the EFS- (Figure 8). Furthermore, primary relaxation re- induced primary relaxation above and beyond the sponses with a combination of 10 nM phentolamine time control at all frequencies tested (Figures 6 and and 30 nM sildena®l were quantitatively similar to 7). Additionally, the contractile responses were the responses obtained in the presence of 50 nM consistently attenuated in the presence of 20 nM sildena®l alone. phentolamine, reversing the trend of increased When NANC relaxation responses were isolated contraction at high frequencies ( 15 Hz) in the by treatment with bretylium and atropine, sildena®l control (Figures 5 and 7). Sildena®l also potentiated caused an enhancement of EFS-induced NANC

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 33

Figure 8 Comparison of combined treatment with phentolamine and sildena®l versus sildena®l alone on nerve-mediated relaxation in rabbit corpus cavernosum. Responses are expressed as `percent change', relative to control. See Figure 6 for experimental details.

to differentiate these two types of interactions by subtracting the initial effect of a given drug prior to the addition of the second compound. After this type of normalization analysis was performed, the enhancement by phentolamine persisted for VIP and PGE1-induced relaxation but not for sildena®l-induced relaxation. This suggests a potential synergistic interaction between antagonism of alpha-adrenergic signaling and cAMP-mediated relaxation, whereas the interaction between phentolamine and cGMP-mediated relaxation appears to be additive. Figure 7 Summary of biphasic responses to electrical ®eld In examining the effect of sildena®l, VIP or PGE1 stimulation for rabbit cavemosal tissue in the absence or presence on phentolamine-induced relaxation, normalization of phentolamine. Contractile (A) and relaxatory (B) responses of dose response data led to similar conclusions for were quanti®ed independently and expressed as `percent change', sildena®l (additive effect) and PGE1 (synergistic relative to control. Attenuated responses are expressed as negative values while potentiated responses are expressed as effect), but not for VIP. Phentolamine-induced positive values. See Figure 6 for experimental details. relaxation was enhanced by VIP, but this potentiation did not persist when the data were normalized. This disparity between the effect of VIP on phento- relaxation at frequencies 2 Hz whereas phentol- lamine-induced relaxation versus the effect of amine did not change the responses, relative to the phentolamine on VIP-induced relaxation may be time control (data not shown). The combination of due, in part, to the relative predominance of the two phentolamine and sildena®l produced responses different receptor-mediated signalling pathways. that were similar to sildena®l alone. Blockade of alpha-adrenergic receptor-mediated contraction by phentolamine may be more effective in facilitating relaxation than VIP in penile trabe- Discussion cular smooth muscle. If this were the case, phentolamine-induced relaxation would overcome any slight synergism produced by the addition of VIP. It Our results indicate that blockade of alpha-adren- is interesting to note that phentolamine enhance- ergic receptors with phentolamine increases the ment of the VIP response was greater in penile ef®cacy of sildena®l, VIP or PGE1 to induce smooth cavernosal tissue from men with erectile dysfunc- muscle relaxation. These vasodilators also enhanced tion than in penile cavernosal tissue from healthy phentolamine-induced relaxation. Since phentola- rabbits. It is unclear if this difference arises from mine and cyclic nucleotide-dependent vasodilators species variation or from compromised sensitivity to cause relaxation by different intracellular mechan- VIP, associated with erectile dysfunction. Differ- isms, enhanced relaxation may be attributed to ences may also exist between EP and VIP receptor- either additive or synergistic effects. We attempted mediated stimulation of adenylate cyclase,32 but

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 34 characterization of these interactions requires The sequence of responses during electrical further investigation. stimulation provide some insight into the integrated Thus, our data are consistent with the hypothesis regulation of cavernosal smooth muscle tone. An that the alpha adrenergic signalling pathway inter- initial transient `primary relaxation' is observed acts directly with the cAMP pathway but is largely immediately after onset of the electrical stimulus. independent of the nitric oxide pathway. Since both This is followed by the contractile response which subtypes of alpha adrenergic receptors are present persists during the course of the stimulation and on the surface of penile trabecular smooth muscle immediately ceases upon termination of EFS. Typi- cells,5±7post-junctional alpha-2 adrenergic receptors cally, cavernosal tissue strips exhibit a `secondary are the most likely mediators of the interaction with relaxation' following the termination of EFS. It is the cAMP pathway. Alpha-2 adrenergic receptors likely that multiple factors mediate this prolonged have been shown to inhibit cAMP-mediated re- relaxation. The ability of the contractile response to sponses in many different systems. In cultured overcome the initial relaxation suggests that the human and rabbit corpus cavernosum smooth effect of adrenergic nerves predominates over that of muscle cells, PGE1 and forskolin-induced cAMP relaxatory non-adrenergic, non-cholinergic nerves. synthesis was inhibited by the alpha-2 adrenergic Rapid inactivation of the relaxatory response during receptor agonist UK14304.6,20 Accumulation of cAMP stimulation is unlikely since all nerve endings is inhibited by alpha-2 adrenergic receptor agon- within the tissue are stimulated indiscriminantly. ists in brain, adipose and retinal tissue.33 35 In Furthermore, when cavernosal tissue strips are the rat jejunum, stimulation of alpha-2 adrenergic treated with inhibitors of adrenergic neurotransmis- receptors antagonizes the effects of PGE1 and VIP on sion, only relaxatory responses are observed during intestinal ¯uid secretion.36 Cyclic AMP-mediated the full course of electrical stimulation.30,31 insulin release from pancreatic islet cells is also The dependence of the contractile response on inhibited by alpha-2 adrenergic receptor agonists.37 active stimulation and the presence of the slower In the penile corpus cavernosum, phentolamine `secondary relaxation' in the absence of active antagonizes norepinephrine binding and may relieve stimulation suggests that the relaxatory response is the inhibition of cAMP synthesis by alpha-2 longer lasting and is `unmasked' when adrenergic adrenergic receptors. input ceases. Thus, phentolamine, at suf®cient While evaluating cavernosal tissue responses to concentrations, should attenuate the adrenergic exogenous vasodilators are informative, the interac- input and decrease the `resistance' to relaxation. tion between alpha adrenergic blockade and vaso- In this manner, phentolamine may potentiate the dilators may be further characterized by examining primary relaxation and be more effective at higher neurogenic responses in penile corpus cavernosum frequencies of stimulation where adrenergic input is tissue strips. This experimental paradigm enabled greater. Sildena®l alone also enhanced the primary us to study the action of phentolamine on endogen- relaxation, presumably by stabilizing intracellular ous vasoactive substances which are released during cGMP synthesized in response to nitric oxide. The electrical stimulation. Since tissue strips were combined treatment of phentolamine and a lower contracted with endothelin-1, a non-adrenergic dose of sildena®l produced similar relaxation to a vasoconstrictor which is not directly antagonized higher dose of sildena®l alone. For instance, the by phentolamine, the only source of adrenergic relaxation response to EFS with a combination of input was norepinephrine released from nerve 10 nM phentolamine and 30 nM sildena®l was endings in the cavernosal tissue during electrical similar to that obtained with 50 nM sildena®l alone. stimulation. We observed that phentolamine mod- In the presence of adrenergic and cholinergic i®ed the biphasic response to electrical stimulation blockade, the contractile response in cavernosal by potentiating relaxation and attenuating contrac- tissue strips is abolished and only relaxation is tion. These ®ndings suggest that antagonism of observed during electrical stimulation. Thus, if alpha-adrenergic neurogenic input enables other phentolamine enhances relaxation by attenuating independent relaxatory pathways to predominate contraction mediated by alpha adrenergic receptors, within penile trabecular smooth muscle. In addition one would expect to observe no effect by a pure to nitric oxide, it is likely that VIP is also released as adrenergic antagonist on isolated NANC responses. a relaxatory co-transmitter. The release of prosta- Indeed, phentolamine did not signi®cantly enhance glandin E during electrical stimulation of isolated EFS-induced NANC relaxation relative to the time cavernosal tissue strips has not been reported. control. In contrast, sildena®l remained ef®cacious However, in the intact penis, increased blood ¯ow in potentiating EFS-induced NANC relaxation. This and tissue oxygenation, secondary to NANC neuro- is not surprising, since the nitric oxide pathway transmitter release, may stimulate the production predominates in the mediation of NANC relaxa- and release of prostanoids from endothelial and tion.14 A secondary mechanism of action has also smooth muscle cells.38±40 The involvement of other been reported for phentolamine. In addition to the relaxatory neurotransmitters or paracrine=autocrine direct antagonism of catecholamine action on alpha- substances remains to be de®ned. 1 and alpha-2 adrenergic receptors, phentolamine

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 35 erection. This mechanism of action may substantiate the use of phentolamine as an adjunct therapy for erectile dysfunction when used in combination with other drugs which initiate vasodilation through different mechanisms.

Acknowledgements

We would like to thank Ms Cynthia Gallant for her efforts in executing some of the experiments and Ms Jerie McGrath-Cerqua for her administrative assistance. This work was supported by grants R01-DK39080, R01-DK40025, R29-DK47950 and K01-DK02696 from the National Institute of Diabetes and Digestive and Kidney Diseases.

References

1 Adaikan PG, Karim SM. Adrenoreceptors in the human penis. Auton Pharmacol 1981; 1: 199±203. 2 Hedlund H, Andersson K-E. Comparison of the responses to drugs acting on adrenoreceptors and muscarinic receptors in human isolated corpus cavernosum and cavernous artery. Auton Pharmacol 1985; 5: 81±88. 3 Levin RM, Wein AJ. Adrenergic alpha receptors outnumber beta receptors in human penile corpus cavernosum. Invest Urol 1980; 18: 225±226. 4 Saenz de Tejada I, et al. Regulation of adrenergic activity in penile corpus cavernosum. J Urol 1989; 142: 1117±1121. 5 Traish AM, Moreland RB, Huang Y-H, Goldstein I. Expression Figure 9 Proposed functional interactions between the alpha- of functional alpha2-adrenergic receptor subtypes in human adrenergic pathway and cyclic nucleotide mediated pathways in corpus cavernosum and in cultured trabecular smooth muscle penile corpus cavernosum smooth muscle. cells. Recept Signal Transduct 1997; 7: 55±67. 6 Gupta S, et al. The expression of functional postsynaptic alpha2-adrenoceptors in the corpus cavernosum smooth muscle. Br J Pharmacol 1998; 123: 1237±1245. may act on the endothelium to stimulate nitric oxide 7 Traish AM, et al. Identi®cation of alpha 1-adrenergic receptor synthase or increase the ef®cacy of nitric oxide subtypes in human corpus cavernosum tissue and in cultured released from the endothelium.41 This effect was not trabecular smooth muscle cells. Receptor 1995; 5: 145±157. apparent in the present study and the signi®cance of 8 Cotecchia S, et al. Multiple second messenger pathways of alpha-adrenergic receptor subtypes expressed in eukaryotic this alternative pathway remains unclear. cells. J Biol Chem 1990; 265: 63±69. In summary, we observe enhancement of penile 9 Exton JH. Mechanisms involved in alpha-adrenergic pheno- cavernosal smooth muscle relaxation with phentol- mena. Am J Physiol 1985; 248: E633±E647. amine only when adrenergic input is present. While 10 MacKinnon AC, Speding M, Brown CM. a2-adrenoceptors: more subtypes but fewer functional differences. Trends other constrictive factors such as endothelins and Pharmacol Sci 1994; 15: 119±123. prostanoids may be important in maintaining ¯ac- 11 Brindley GS. Cavernosal alpha blockade: a new technique for cidity, it is reasonable to assume that adrenergic investigating and treating erectile impotence. Br J Psychiatry tone may also predominate over relaxatory mechan- 1983; 143: 332±337. isms in the ¯accid state of the penis. Thus, 12 Diederichs W, Stief CG, Lue TF, Tanagho EA. Norepinephrine involvement in penile detumescence. J Urol 1990; 143: 1264± phentolamine could reduce the predominance of 1266. the contractile pathway and the `resistance to 13 Blum MD, Bahnson RR, Porter TN, Carter MF. Effect of local relaxation' (see Figure 9). This may prove to be alpha-adrenergic blockade on human penile erection. J Urol important in cases where basal adrenergic tone is 1985; 134: 479±481. 14 Andersson K-E, Wagner G. Physiology of penile erection. greater than normal. The responsiveness of human Physiol Rev 1995; 75: 191±236. erectile tissue to alpha adrenergic agonists is 15 Murad F, et al. Guanylate cyclase: activation by azide, nitro thought to be altered by age and disease and the compounds, nitric oxide, and hydroxyl radical and inhibition ef®cacy of phenylephrine-induced contraction is by hemoglobin and myoglobin. Adv Cyclic Nucleotide Res signi®cantly increased in cavernosal tissue from 1978; 9:145±158. 42±44 16 Pierce KL, Gil DW, Woodward DF, Regan JW. Cloning of impotent men. Thus, it is possible that adren- human prostanoid receptors. Trends Pharmacol Sci 1995; 16: ergic imbalance towards vasoconstriction impairs 253±256.

International Journal of Impotence Research Phentolamine and vasodilators NN Kim et al 36 17 Pierce KL, Regan JW. Prostanoid receptor heterogeneity 32 Beubler E. VIP and PGE1 activate adenylate cyclase in rat through alternative mRNA splicing. Life Sci 1998; 62: 1479± intestinal epithelial cell membranes via different mechanisms. 1483. Eur J Pharmacol 1981; 74: 67±72. 18 Woodward DF, Regan JW, Lake S, Ocklind A. The molecular 33 Duman RS, Enna SJ. A procedure for measuring alpha biology and ocular distribution of prostanoid receptors. Surv 2-adrenergic receptor-mediated inhibition of cyclic Ophthalmol 1997; 41 (Suppl 2): S15±S21. AMP accumulation in rat brain slices. Brain Res 1986; 384: 19 Breyer MD, Jacobson HR, Breyer RM. Functional and 391±394. molecular aspects of renal prostaglandin receptors. JAm 34 Garcia-Sainz JA, Martinez OM. Functional alpha 2-adreno- Society Nephrol 1996; 7: 8±17. ceptors in rat adipocytes: inhibition of cyclic AMP accumula- 20 Traish AM, Moreland RB, Huang Y-H, Goldstein I. G-protein- tion. Biochem Int 1989; 19: 899±907. coupled receptor agonists augment adenylyl cyclase activity 35 Osborne NN. Inhibition of cAMP production by alpha 2- induced by forskolin in human corpus cavernosum smooth adrenoceptor stimulation in rabbit retina. Brain Res 1991; 553: muscle cells. Recept Signal Transduct 1997; 7:121±132. 84±88. 21 Sugimoto Y, et al. Two isoforms of the EP3 receptor with 36 Nakaki T, Nakadate T, Yamamoto S, Kato R. Alpha-2 different carboxyl-terminal domains. Identical ligand binding adrenergic inhibition of intestinal secretion induced by properties and different coupling properties with Gi proteins. J prostaglandin E1, vasoactive intestinal peptide and dibutyryl Biol Chem 1993; 268: 2712±2718. cyclic AMP in rat jejunum. J Pharmacol Exp Ther 1982; 21: 22 Namba T, et al. Alternative splicing of C-terminal tail of 648±653. prostaglandin E receptor subtype EP3 determines G-protein 37 Nakaki T, Nakadate T, Yamamoto S, Kato R. Inhibition speci®city. Nature 1993; 365: 166±170. of dibutyryl cyclic AMP-induced insulin release by 23 Irie A, et al. Third isoform of the prostaglandin-E-receptor EP3 alpha-2 adrenergic stimulation. Life Sci 1983; 32: 191± subtype with different C-terminal tail coupling to both 195. stimulation and inhibition of adenylate cyclase. Eur J Biochem 38 Daley JT, et al. Prostanoid production in rabbit corpus 1993; 217: 313±318. cavernosum: I. regulation by oxygen tension. J Urol 1996; 24 Azadzoi KM, Payton T, Krane RJ, Goldstein I. Effects of 155: 1482±1487. intracavernosal trazodone hydrochloride: animal and human 39 Moreland RB, et al. Oxygen tension modulates TGF-b1 studies. J Urol 1990; 144: 1277±1282. expression and PGE production in human corpus cavernosum 25 Abber JC, et al. Priapism induced by chlorpromazine smooth muscle cells. Mol Urol 1998; 2: 41±47. and trazodone: mechanism of action. J Urol 1987; 135: 1039± 40 Shaul PW, Campbell WB, Farrar MA, Magness RR. Oxygen 1042. modulates prostacyclin synthesis in ovine fetal pulmonary 26 Gwinup G. Oral phentolamine in nonspeci®c erectile insuf®- arteries by an effect on cyclooxygenase. J Clin Invest 1992; 90: ciency. Ann Int Med 1988; 109: 162±163. 2147±2155. 27 Saenz de Tejada I, et al. Pathophysiology of prolonged 41 Traish A, et al. Phentolamine mesylate relaxes penile corpus penile erection associated with trazodone use. JUrol1991; cavernosum tissue by adrenergic and non-adrenergic mechan- 145: 60±64. isms. Int J Impot Res 1998; 10: 215±223. 28 Zorgniotti AW. Experience with buccal phentolamine mesy- 42 Dahiya R, et al. mRNA and protein expression of nitric oxide late for impotence. Int J Impotence Res 1994; 6: 37±41. synthase and adrenoceptor alpha 1 in young and old rat penile 29 Kaplan SA, et al. Combination therapy using oral alpha- tissues. Br J Urol 1997; 80: 300±306. blockers and intracavernosal injection in men with erectile 43 Christ GJ, Maayani S, Valcic M, Melman A. Pharmacological dysfunction. Urology 1998; 52: 739±743. studies of human erectile tissue: characteristics of spon- 30 Kim N, Azadzoi KM, Goldstein I, Saenz de Tejada I. A nitric taneous contractions and alterations in alpha-adrenoceptor oxide-like factor mediate nonadrenergic-noncholinergic neu- responsiveness with age and disease in isolated tissues. Br J rogenic relaxation of penile corpus cavernosum smooth Pharmacol 1990; 101: 375±381. muscle. J Clin Invest 1991; 88: 112±118. 44 Christ GJ, Stone B, Melman A. Age-dependent alterations in 31 Saenz de Tejada I, et al. Cholinergic neurotransmission in the ef®cacy of phenylephrine-induced contractions in vascu- human corpus cavernosum. I. Responses of isolated tissue. Am lar smooth muscle isolated from the corpus cavernosum of J Physiol 1988; 254: H459±H467. impotent men. Can J Physiol Pharmacol 1991; 69: 909±913.

International Journal of Impotence Research