Pharmacological characterization of the relaxant effect induced by adrenomedullin in rat cavernosal smooth muscle

The aim of the present study was to determine the mechanisms underlying the relaxant effect of adrenomedullin (AM) in rat cavernosal smooth muscle (CSM) and the expression of AM system components in this tissue. Functional assays using standard muscle bath procedures were performed in CSM isolated from male Wistar rats. Protein and mRNA levels of pre-pro-AM, calcitonin receptor-like receptor (CRLR), and Subtypes 1, 2 and 3 of the receptor activity-modifying protein (RAMP) family were assessed by Western immunoblotting and quantitative real-time polymerase chain reaction, respectively. Nitrate and 6-keto-prostaglandin F1α (6-keto-PGF1α; a stable product of prostacyclin) levels were determined using commercially available kits. Protein and mRNA of AM, CRLR, and RAMP 1, -2, and -3 were detected in rat CSM. Immunohistochemical assays demonstrated that AM and CRLR were expressed in rat CSM. AM relaxed CSM strips in a concentration-dependent manner. AM22-52, a selective antagonist for AM receptors, reduced the relaxation induced by AM. Conversely, CGRP8-37, a selective antagonist for calcitonin gene-related peptide receptors, did not affect AM-induced relaxation. Preincubation of CSM strips with NG-nitro-L-arginine-methyl-ester (L-NAME, nitric oxide synthase inhibitor), 1H-(1,2,4)oxadiazolo[4,3-a]quinoxalin-1-one (ODQ, quanylyl cyclase inhibitor), Rp-8-Br-PET-cGMPS (cGMP-dependent protein kinase inhibitor), SC560 [5-(4-chlorophenyl)-1-(4-methoxyphenyl)-3-trifluoromethyl pyrazole, selective cyclooxygenase-1 inhibitor], and 4-aminopyridine (voltage-dependent K+ channel blocker) reduced AM-induced relaxation. On the other hand, 7-nitroindazole (selective neuronal nitric oxide synthase inhibitor), wortmannin (phosphatidylinositol 3-kinase inhibitor), H89 (protein kinase A inhibitor), SQ22536 [9-(tetrahydro-2-furanyl)-9H-purin-6-amine, adenylate cyclase inhibitor], glibenclamide (selective blocker of ATP-sensitive K+ channels), and apamin (Ca2+-activated channel blocker) did not affect AM-induced relaxation. AM increased nitrate levels and 6-keto-PGF1α in rat CSM. The major new contribution of this research is that it demonstrated expression of AM and its receptor in rat CSM. Moreover, we provided evidence that AM-induced relaxation in this tissue is mediated by AM receptors by a mechanism that involves the nitric oxide-cGMP pathway, a vasodilator prostanoid, and the opening of voltage-dependent K+ channels.


Introduction
The initiation and maintenance of penile erection is caused by relaxation of the blood vessels in the cavernosal smooth muscle (CSM), which results in an increased blood flow into the trabecular spaces of the corpora cavernosa (1). The control of CSM tone is mainly mediated by the adrenergic, cholinergic, and nonadrenergic, noncholinergic (NANC) systems (2). Noradrenaline, released from sympathetic nerves, induces contraction of penile CSM while parasympathetic innervations mediate smooth muscle relaxation in the trabecular network and cavernosal arterial venous bed (3). Nitric oxide (NO) released from NANC nerve endings and from the vascular endothelium is considered the most important mediator of CSM relaxation (4). Release of endothelium-derived NO is regulated by several factors, including vasodilator substances such as adrenomedullin (AM), which has been demonstrated to play a role as a modulator of erectile function (5)(6)(7).
AM consists of a 52-amino acid peptide, initially isolated from human pheochromocytoma cells, that displays vasorelaxant and hypotensive actions (8). AM has a ring structure formed by a disulfide bond and an amidated carboxyl terminal, and belongs to a family of peptides that include amylin and calcitonin gene-related peptide (CGRP) (8). In the vasculature, the relaxant response induced by AM is mediated by the seven-transmembrane G protein-coupled calcitonin receptor-like receptor (CRLR), which coassembles with Subtypes 2 and 3 of the receptor activity-modifying protein (RAMP) family, thus forming a receptor-coreceptor system (9,10). Although the vasodilator effect of AM in different blood vessels is well characterized (10), few reports have described the effect of AM in CSM relaxation. However, it has been reported that intracavernosal injections of AM increased cavernosal pressure and penile length in cats (5). This response was not mediated by CGRP receptors and did not involve NO generation or the opening of K + channels (5,6). In anesthetized rats, intracavernosal administration of AM resulted in increased cavernous pressure and penile erection, which was attenuated by inhibitors of the NO-cGMP pathway (7). The relaxation induced by AM in isolated rabbit CSM strips does not involve NO, vasodilator prostanoids, or the opening of K + channels (11). Finally, AM is localized in human endothelial cells of cavernous vessels, where it may contribute to penile erection (12). These findings imply that AM is a modulator of CSM tone and suggest that AM might potentiate erectile function. Moreover, based on the above-mentioned observations, it is possible to conclude that the mechanism by which AM induces vasorelaxation or erection varies with species, vascular bed studied, and experimental procedure employed.
The AM system has been postulated to have a cardioprotective role in a wide range of diseases (13). Cardiovascular diseases are often associated with erectile dysfunction (ED) (14), and, in this case, increased levels of AM may play a compensatory role for ED. Isolated CSM is a useful model for the study of penile erectile responses and ED (15,16). Thus, the study of physiological expression and function of AM receptors in CSM may provide valuable information on the contribution of AM to CSM tone. The effect of AM on cavernous pressure and penile erection has been previously evaluated in anesthetized rats using intracavernous pressure measurements (7). However, to the best of our knowledge, there are no reports describing the receptors involved in AM-induced relaxation of rat CSM or the detailed mechanisms underlying such a response. The aims of the present study were to attempt a functional characterization of the AM receptors in rat CSM and to investigate the mechanisms underlying AM-induced relaxation in this tissue. In addition, quantitative real-time polymerase chain reaction (qRT-PCR), Western immunoblotting, and immunohistochemical assays were performed to verify expression of AM, CRLR, and RAMP1, -2, and -3 in rat CSM.

Animals
Male Wistar rats weighing 250-300 g (50-70 days of age) were housed under standard laboratory conditions with free access to food and water. The housing conditions and experimental protocols were approved by the Animal Ethics Committee of the Universidade de Sa˜o Paulo, Campus of Ribeira˜o Preto, Brazil (Protocol #10. 1.1293.53.4). The animals were anesthetized with isoflurane [2-chloro-2-(difluoromethoxy)-1,1,1-trifluoroethane] and killed by aortic exsanguination. CSM was removed for functional assays, Western immunoblotting, qRT-PCR, and immunohistochemical experiments.

qRT-PCR
Total cellular RNA was extracted using Trizol 1 Reagent (Invitrogen, USA), and RNA was reverse transcribed to single-stranded cDNA using a High Capacity Kit (Applied Biosystems, USA) according to the manufacturer's protocol. For quantitative analysis of the genes of interest [pre-pro-AM (Rn 00562327_m1), CRLR (Rn 00562334_m1), RAMP1 (Rn 01427056_m1), RAMP2 (Rn 00824652_m1), and RAMP3 (Rn 00571815_m1)], a commercially available TaqMan Assay-on-Demand System that consists of a kit of oligonucleotides and probes was used (Applied Biosystems). Reverse transcription was performed using 1 mg total RNA for each sample in 20 mL of the total reaction mixture. The cDNA obtained was diluted 1:10, and 4.5 mL was used for each 10 mL of the qRT-PCR mixture using the TaqMan Master Mix (Applied Biosystems). Reactions were carried out in duplicate and analyzed with 7500 Sequence Detection System apparatus (Applied Biosystems). Data were analyzed using the ABI-7500 SDS software (Applied Biosystems). Total RNA absorbed was normalized on the basis of the Ct value for the GAPDH gene (Rn 01775763_m1). The variation in expression among samples was calculated by the 2 -DDCt method, and the mean delta Ct value for a group of six samples from the control was used for calibration (17).

Immunohistochemistry
Paraffin-embedded CSM segments were stained using the avidin-biotinylated peroxidase complex method. Briefly, 4-mm sections (Reichert Jung 2040 microtome, Leica, USA) were cut, deparaffinized with xylene and dehydrated in ethanol. Endogenous peroxidase and biotin were blocked by immersing slides in 3% hydrogen peroxide. The sections were incubated with the following primary antibodies: AM (sc-16496, 1:250) and CRLR (sc-18007, 1:250). The reactions were revealed using 0.2 mg/mL diaminobenzidine solution (10 mg tablets in 50 mL PBS 0.01 M, pH 7.4; D5905; Sigma-Aldrich, USA) and stained by Harris hematoxylin. On each slide, two fields were selected in areas with high concentrations of positive cells or stained cells, using 506 or 10006 magnification. The slides were analyzed using a Leica model DM 5500 B microscope. The images were registered using a Leica digital camera DFC 290 (3MP) attached to the microscope and filed using the Leica QWin software.
Functional studies CSM was isolated as described previously (16). In brief, the penis was harvested by cutting the corporeal body at the level of its attachment to the ischium bone and then immersed in Krebs solution (130 mM NaCl, 4.7 mM KCl, 1.18 mM KH 2 PO 4 , 1.17 mM MgSO 4 .7H 2 O, 1.6 mM CaCl 2 .2H 2 O, 14.9 mM NaHCO 3 , and 5.5 mM glucose). The tunica albuginea was carefully opened from its proximal extremity toward the penile shaft, and the erectile tissue within the corpus cavernosum was surgically dissected free. Strips of CSM (161610 mm) were mounted in a 5-mL organ chamber containing Krebs solution at 376C and continuously bubbled with a gas mixture of 95% oxygen and 5% carbon dioxide, pH 7.4. One end of each corporal strip was attached to the bottom of the organ bath and the other end was tied to a force transducer (TRI201, Panlab, Spain). The strips were stretched to a resting tension of 3 mN and allowed to equilibrate for 60 min. The responses were recorded on a computer system using Chart Pro 5 (PowerLab, ADInstruments, Australia).
CSM strips were precontracted with phenylephrine (10 mM), and when the contraction reached a plateau, concentration-response curves for AM (10 fM to 30 nM) were obtained by stepwise increase of the agonist concentration. Additions were made as soon as a steady response was obtained from the preceding concentration. For comparison, concentration-response curves for CGRP (1 pM to 0.3 mM) and acetylcholine (1 nM to 1 mM) were also obtained in precontracted CSM strips. Relaxation is reported as the percent change from phenylephrine-contracted levels.

Nitrate measurements
Nitrate (NO 3 -, a metabolite of NO) levels were measured in supernatants from CSM homogenates. The strips were contracted with 10 mM phenylephrine and then exposed to 30 nM AM or 100 mM L-NAME. Some strips were incubated with 100 mM L-NAME for 30 min prior to the administration of AM. When the maximal relaxation induced by AM was achieved, tissues were frozen in liquid nitrogen. CSM was homogenized in 200 mL PBS buffer, pH 7.4, and centrifuged at 10,000 g (10 min, 46C). The supernatant was ultrafiltered (Amicon Ultra-0.5 mL 10 kDa, Millipore, USA) at 14,000 g (15 min, 256C). A commercially available kit (#780001, Cayman, USA) was used to measure nitrate levels. Results are reported as mM/ mg protein. Protein concentrations were determined with a protein assay reagent (Bio-Rad Laboratories, USA).
The strips were contracted with 10 mM phenylephrine and were then exposed to 30 nM AM. When the maximal relaxation induced by AM was achieved, the strips were frozen in liquid nitrogen. CSM was homogenized in EIA buffer (1 M phosphate solution containing 1% BSA, 4 M sodium chloride, 10 mM EDTA and 0.1% sodium azide) and centrifuged at 2000 g (15 min, 46C). The samples (50 mL) were deproteinized by precipitation using 50 mL absolute ethanol kept at 46C, followed by stirring and then kept for 30 min in a freezer at --206C. The supernatant was centrifuged at 4000 g (10 min, 256C). Levels of 6keto-PGF 1a were measured using a commercially available kit (Cayman, code 515211). Results are reported as pg/mg protein. Protein concentrations were determined with a protein assay reagent (Bio-Rad Laboratories).

Drugs
ODQ, 7-nitroindazole, SC560, and glibenclamide were prepared as stock solutions in dimethyl sulfoxide (DMSO), whereas the other drugs were dissolved in distilled water. The bath concentration of DMSO did not exceed 0.5%, which was shown to have no effects per se on basal tonus of the preparations or on agonist-mediated relaxation.

Statistical analysis
Data are reported as means±SE. Statistically significant differences were determined by the Student t-test or analysis of variance (ANOVA) followed by the Bonferroni multiple comparison test. P,0.05 was considered to be statistically significant.

Results
Protein and mRNA expression of AM system components in rat CSM Figure 1A shows representative immunoblots for AM, CRLR, and RAMP1, -2, and -3 protein expression in rat CSM. The results obtained by qRT-PCR showed that rat CSM expressed mRNA of pre-pro-AM, CRLR, and RAMP1, -2, and -3 ( Figure 1B).
Expression and localization of AM and CRLR in rat CSM. Immunohistochemical studies revealed staining for AM and CRLR in rat cavernous tissue. Nuclear staining for both AM and CRLR were detected diffusely in all constituents of the cavernous tissue including connective tissue, in the endothelium lining vascular spaces, and in smooth muscle (Figure 2).

Nitrate and 6-keto-PGF 1a measurements
AM significantly increased 6-keto-PGF 1a (a stable product of PGI 2 ) in rat CSM compared with tissues that were not stimulated with the peptide ( Figure 8A). AM significantly increased nitrate generation in rat CSM compared with tissues that were not stimulated with the peptide ( Figure 8B). AM-induced nitrate generation was significantly inhibited by L-NAME, which had no effect per se on basal nitrate levels.

Discussion
In the present study, protein and mRNA expression of AM, CRLR, and RAMP1, -2, and -3 were detected in rat CSM. Immunohistochemical assays showed that AM and CRLR are expressed in the cavernous tissue. AM acts as a circulating hormone and locally in an autocrine/ paracrine fashion. Because AM is expressed in rat CSM, it may play a role in the autocrine/paracrine regulation of penile erection due to its vasodilator action. AM is considered an important regulatory peptide that helps to regulate cardiovascular homeostasis. AM levels in cardiovascular tissues are elevated to compensate for changes induced by cardiovascular diseases such as atherosclerosis and hypertension (24). Thus, increased AM expression in CSM could exert a protective action against ED. In fact, it has been suggested that combination therapy using PGE 1 and proerection agents such as AM may be beneficial in the treatment of ED (25).
A pharmacological characterization of the mechanisms mediating the relaxant effect of AM in rat CSM was attempted with functional assays, using standard muscle bath procedures. AM induced CSM relaxation in a concentration-dependent manner. AM was similar in potency to CGRP, and both were more potent than acetylcholine, which is in accordance with previous findings in rat aorta (26), rat mesenteric arterial bed (27), and cat CSM (6). Relaxation induced by AM has Figure 6. Relaxation responses induced by adrenomedullin (AM) on rat cavernosal smooth muscle strips pre-contracted with phenylephrine. The concentration-response curves were obtained in the absence (control) or after incubation for 30 min with the following drugs: 100 mM L-NAME, 100 mM 7-nitroindazole, 1 mM ODQ, 3 mM Rp-8-Br-PET-cGMPS, 10 mM sildenafil, 1 mM wortmannin, 10 mM SC560, or the combination of L-NAME and SC560. Data are reported as means±SE of 5 to 6 independent preparations. been previously described in isolated rabbit CSM in a concentration range different from that employed in the present study (11). A possible explanation for such discrepancy is that the mechanism by which AM induces vasorelaxation or erection varies with species, vascular bed studied, and experimental procedure employed (5)(6)(7)11,28).
The AM receptor is composed of the CRLR and specific RAMP (9,10). RAMPs are a class of type I transmembrane proteins that interact with and modulate the activities of G protein-coupled receptors. Cell surface RAMP2-CRLR and RAMP3-CRLR complexes are AM receptors, while the RAMP1-CRLR complex forms the CGRP receptor (9,10). RAMP interaction with its associated receptor can lead to three potential consequences: trafficking of receptor protein from an intracellular compartment to the cell surface, alteration in the terminal glycosylation of the receptor, and alteration of receptor phenotype, presumably through a direct or indirect effect on the ligand-binding site (29). Although the antagonist AM  has been shown to selectively inhibit AM receptors, CGRP 8-37 is a CGRP receptor antagonist that has been shown to be able to block some, but not all, of the actions of AM in the vasculature (30). This observation indicates that the vasorelaxation induced by AM may occur due to its interaction with both AM or CGRP receptors. The present findings show that AM-induced CSM relaxation was attenuated by AM 22-52 , but not CGRP  . Our study provides the first functional evidence that relaxation induced by AM in rat CSM is solely mediated by AM receptors.
Activation of adenylate cyclase with consequent increase in cAMP and cAMP-dependent protein kinase activation has been implicated in the vascular relaxation induced by AM (31,32). In our study, neither SQ22536 nor H89 altered AM-induced relaxation, which is not consistent with the participation of adenylate cyclase and protein kinase in this response.  In some vascular tissues, AM induces relaxation via production of NO, with consequent increases in cGMP levels (33,34). NO is formed from L-arginine by the catalytic action of the enzyme NOS. The latter has three isoforms: nNOS (or NOS type I), inducible NOS (or NOS type II), and endothelial NOS (eNOS or NOS type III). nNOS and eNOS are the main isoforms involved in penile erection and are present in the nerves and endothelium of the penis, respectively (35). Our data show that L-NAME partially, but significantly, reduced AM-mediated relaxation. In addition, AM increased nitrate levels in rat CSM, and this response was inhibited by L-NAME, further implicating NOS in this process. Taken together, these results show that activation of NOS with consequent NO generation play a role in AM-mediated relaxation. 7-Nitroindazole, a selective nNOS inhibitor, had no effect on AM-induced relaxation, suggesting that this NOS isoform could not account for the AM-mediated relaxation in rat CSM. The selective inhibitor of guanylyl cyclase enzyme, ODQ, reduced the relaxant action of AM, confirming the involvement of the NO-cGMP pathway in this response as previously observed in cat and rat CSM (5-7). cGMPstimulated protein kinase (PKG) acts downstream to reduce Ca 2+ + concentration and/or the sensitivity of the contractile proteins to Ca 2+ + , thus leading to smooth muscle relaxation. Our findings show that PKG activation plays a role in AM-induced relaxation because Rp-8-Br-PET-cGMPS reduced this response. Phosphodiesterase type 5 (PDE5) is widely expressed in CSM, where it catalyzes cGMP hydrolysis (36). Sildenafil, a PDE5 inhibitor, induced a leftward displacement of the concentration-response curves for AM, further suggesting that PDE5 negatively modulates the relaxation induced by AM.
AM has also been shown to elicit phosphatidylinositol 3-kinase (PI3K) activation and Akt phosphorylation, resulting in the stimulation of eNOS (37). Our findings with wortmannin discard the participation of the PI3K/Akt pathway on AM-induced relaxation. It is important to note that blockade of the NO-cGMP pathway only partially attenuated the relaxant response induced by AM, indicating that mediators unrelated to the production of NO also participate in this response. In fact, we observed that the relaxation evoked by AM was partially blunted by SC560, an inhibitor of cyclooxygenase-1, suggesting the involvement of vasodilator prostanoids in the relaxing effect of AM. Additionally, AM increased 6-keto-PGF 1a , a stable product of PGI 2 . These results agree with a previous finding showing the participation of prostanoids in AMinduced relaxation in porcine ciliary arteries (38). When L-NAME and SC560 were simultaneously added to the organ bath, an additional inhibitory effect on AM-induced relaxation was observed, indicating that both NO and vasodilator prostanoids participate in this response.
Activation of K + channels is an important mechanism in vascular smooth muscle hyperpolarization and relaxation, and cGMP can modulate the activity of K + channels to elicit vasodilatation. A role for K + channels in AMmediated relaxation has already been described for vasculature (27,38). We found that 4-aminopyridine, but not glibenclamide or apamin, reduced AM-induced relaxation, indicating that the activation of voltage-sensitive K + channels plays a role in such responses.
The major new finding of the present study is that AM receptors mediate CSM relaxation via the NO-cGMP pathway, vasodilator prostanoids (probably PGI 2 ), and the opening of K + channels. Studies of the expression and function of AM receptors in CSM may provide valuable information on the contribution of AM to CSM tone, since this tissue is a useful model for the study of penile erectile responses and ED. Figure 8. 6-keto-PGF 1a and nitrate levels in rat cavernosal smooth muscle strips stimulated with adrenomedullin (AM). Data are reported as means±SE of 5 to 6 independent preparations. *P,0.05, compared to basal (Student t-test); # P,0.05, compared to basal, L-NAME, and AM+ +L-NAME (ANOVA followed by the Bonferroni multiple comparison test).