0022-534 7/92/14 71-0220$03.00/0 THE JOURNAL OF UROLOGY Copyright© 1992 by AMERICAN UROLOGICAL ASSOCIATION, INC.

Vol. 147, 220-225, January 1992

Printed in U.S.A.

ENDOTHELIUM-DERIVED NITRIC OXIDE AND CYCLOOXYGENASE PRODUCTS MODULATE CORPUS CAVERNOSUM SMOOTH MUSCLE TONE KAZEM M. AZADZOI,* NOEL KIM, MICHAEL L. BROWN, IRWIN GOLDSTEIN, RICHARD A. COHEN AND INIGO SAENZ DE TEJADA From the Departments of Urology and Medicine, Boston University Medical Center and the Boston Veterans Administration Medical Center, Boston, Massachusetts

ABSTRACT

Relaxation of penile corpus cavernosum smooth muscle is controlled by nerve and endothelium derived substances. In this study, endothelium-dependent relaxation of corporal smooth muscle was characterized and the role of arachidonic acid products of cyclooxygenase in endothelium-dependent relaxation was examined. Endothelium removal from rabbit corpora was performed by infusion with 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane sulfonate and was confirmed by transmission electron microscopy. Strips of human and rabbit corporal tissues were studied in the organ chambers for isometric tension measurement. The accumulation of cyclic guanosine monophosphate (cGMP) and the release of eicosanoids from corporal tissue was measured by radioimmunoassay and correlated to smooth muscle relaxation. Our study showed that relaxation of corpus cavernosum tissue to acetylcholine, bradykinin and substance P was endothelium-dependent; potentiated by indomethacin; and inhibited by NG-monomethyl-L-arginine, methylene blue or LY83583. Relaxation to papaverine and sodium nitroprusside was endothelium-independent, and unaffected by NGmonomethyl-L-arginine. Relaxation to vasoactive intestinal polypeptide was partially endotheliumdependent; potentiated by indomethacin; attenuated by NG-monomethyl-L-arginine or methylene blue. The tissue level of cGMP was enhanced by acetylcholine and nitric oxide. Methylene blue inhibited both basal and drug-stimulated levels of cGMP. The release of eicosanoids was enhanced by acetylcholine and blocked by indomethacin. In conclusion, nitric oxide or a closely related substance accounts for the activity of endothelium-derived relaxing factor in the corporal tissue. Inhibition of the release of eicosanoids potentiates the relaxing effect of nitric oxide. Nitric oxide in~reases tissue cGMP which appears to modulate corporal smooth muscle relaxation. KEY WORDS:

muscle, smooth; neuroregulators

Relaxation of corporal smooth muscle is required for penile erection as it allows for the expansion of the lacunar spaces and reduces cavernosal venous outflow by compression of ven ules against the tunica albuginea, the surrounding fibrous structure. Corporal smooth muscle relaxation is under the control of dilator nerves (cholinergic and nonadrenergic noncholinergic) and of the vascular endothelium which lines the lacunar spaces. Vasoactive intestinal polypeptide has been proposed as the nonadrenergic noncholinergic neurotransmitter of dilator nerves in penile smooth muscle due to the presence of immunoreactive VIP within penile nerves and the dilator effects of this peptide on penile smooth muscle. However, all the criteria to consider it the nonadrenergic noncholinergic neurotransmitter in the corpus cavernosum smooth muscle have not been fulfilled. 1 Vascular endothelium releases vasoactive substances which can alter the tone of the surrounding smooth muscle and thereby participate in the regulation of hemodynamics in vascular tissues. 2- 10 Endothelium-derived relaxing factor, endothelium-derived hyperpolarizing factor, prostaglandins, and the peptide endothelin have been demonstrated to be released by the vascular endothelium and can influence smooth muscle tone in blood vessels. 2- 10 Accepted for publication July 9, 1991. *Requests for reprints: Department of Urology, 720 Harrison Ave., Suite 606, Boston MA 02118. Supported by Grants DK-39080, DK-40025, DK-40487, DK-39624, HL-31607 and HL-38731 from the National Institutes of Health and a Grant from the Veterans Administration Medical Service.

The relaxation of human corporal smooth muscle to acetylcholine and to bradykinin requires the presence of intact endothelium.6· 7 It may be speculated that the factors derived from the endothelium lining the lacunar spaces have an important physiological role in the relaxation of corporal smooth muscle. The role of endothelium-derived relaxing factor and eicosanoids has been the subject of many studies in blood vessels 2 • 4 • 5 • 9 • 10• 11 while very little is known about their role in the penile corpus cavernosum smooth muscle. We have used isolated strips of human and rabbit corporal tissue to study endothelium-mediated relaxation of corpus cavernosum smooth muscle, The interaction of eicosanoids with endothelium-derived relaxing factor and the chemical nature of corpus cavernosum endothelium-derived relaxing factor were also investigated. MATERIALS AND METHODS

Human tissue. This study was approved by the Institutional Review Board for Human Studies at our medical center. All patients included in this study underwent penile prosthetic surgery for treatment of impotence. In the operating room, biopsies of the corpora cavernosa were immediately placed in chilled (4C) physiologic salt solution and studied within one hour. Rabbit tissue, The New Zealand White rabbit (three to 3.5 kg.) was chosen as the animal model based on the close similarities that have been reported in the reactivity in vitro of human and rabbit corpus cavernosum. 12 The animals were sacrificed by exsanguination following anesthesia by intravenous administration of nembutal and the penis was removed

220

221

CORPUS CAVERNOSUM SMOOTH MUSCLE TONE

en bloc. A ventral incision was made on the right and left corpora, the tunica was dissected and the corpus cavernosum tissue was exposed. The right and left corporal tissue were dissected and subsequently studied in the organ chambers. Each rabbit provided two to four strips of corpus cavernosum tissue that were studied in separate chambers. Drugs. Phenylephrine hydrochloride, acetylcholine chloride, indomethacin and methylene blue were obtained from Sigma Chemical Co, (St. Louis, MO). Bretylium tosylate was a gift from American Critical Care (Mc Graw Park, IL), N°-monomethyl-L-arginine and 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane sulfonate (CHAPS) were purchased from Calbiochem Corp. (La Jolla, CA). Substance P, bradykinin and VIP were obtained from Bachem, Inc. (Torrance, CA). Indomethacin was mixed in 0.1 M Na2COs and 0.1 M phosphate (pH 6.9). Other drugs were mixed in distilled, deionized water and kept at 4C while in solution. To study cumulative concentration responses, drugs were added to the chamber in half-log increments. Drugs were mixed so that, for every concentration, the volume added to the chamber was 100 microliters. To study the effects of antagonist drugs, tissue was exposed to the study drug for 25 to 30 minutes. Parallel, untreated tissues, were utilized as time controls. Dilutions of nitric oxide were prepared as follows: Test tubes were filled with distilled, deionized water and capped with rubber stoppers. The water was degassed under vacuum at room temperature for 30 minutes. The tubes were then placed on ice and bubbled with pure N 2 gas for an additional 15 minutes via long cannulas (18 gauge, three inch) inserted through the rubber stoppers. Additional shorter needles were inserted through the rubber stoppers to relieve pressure and to enable gas flow. While still on ice, one of the tubes was bubbled with nitric oxide gas for 10 minutes to yield a saturated solution of nitric oxide (-1 mM). Serial dilutions were performed by transferring aliquots between capped tubes with a syringe. Stock solutions were used within 10 minutes of preparation and discarded after each use. The concentrations of nitric oxide stock solutions were determined by a colorimetric assay. To verify the concentration of nitric oxide, stock solutions were allowed to equilibrate with room air and diluted with distilled, deionized water. Aliquots (0.6 ml., in duplicate) were added to 0.8 ml. of 1 % sulfanilic acid in four N HCl, followed by 0.6 ml. of 1% N-(1-naphthyl)ethylenediamine. Absorbances of the samples were read in a spectrophotometer at 548 nm. after 15 min. Standard curves were constructed using NaN02 (two to 20 nanomoles). Preparation of tissue in organ chambers. Strips of human or rabbit corpus cavemosum tissue measuring approximately 3 mm. x 3 mm. x 10 mm. were submerged in 25 ml. organ chambers containing Krebs-physiologic solution. The strips were suspended with a wire to a force transducer on one end, and fixed with silk ties to a metallic support on the opposite end. Composition of the physiologic solution was NaCl, 118.3 mM; KCl, 4.7 mM; MgS0 4 , 0.6 mM; KH 2PO., 1.2 mM; CaClz 2.5 mM; NaHC0 3 , 25 mM; CaNa2 EDTA, 0.026 mM; and glucose, 11.1 mM. The solution was gassed with 95% air and 5% CO2, The pH of the solution was 7.4 and the temperature was maintained at 37C. Isometric tension was measured with a force transducer (model FT03; Grass Instruments, Quincy, MA). The corpus cavernosum tissue was stretched incrementally for a period of two hours and the optimal resting isometric tension for contraction was determined. After every three stretches (0.5 gm. tension/stretch), the tissue was contracted with phenylephrine (2 µM). When the amplitude of the contraction was within 10% of the previous contraction, that tension was considered optimal for isometric contraction. Relaxations were studied following contraction with phenylephrine (0.5 to two µM) or norepinephrine (three µM). When corporal strips were contracted for a second time, to study the effect of a drug (indomethacin on acetylcholine responses) the

tension of the second contraction was always matched to the tension of the first contraction. De-endothelialization of the corporal lacunar spaces. The removal of endothelium from the lacunar spaces of the rabbit corpus cavernosum was performed by a technique previously described for blood vessels with some modification. 13 The intact isolated penis was placed in a tray containing physiologic solution. Twenty-one gauge minicatheters were inserted into the proximal end of the right and left corpora. A third minicatheter was inserted into the distal end of the corpus cavernosum near to the glans penis. After the second and third minicatheters were clamped, three ml. (five mg./ml.) CHAPS was infused into the right corpora through the first minicatheter. After 15 seconds, CHAPS was aspirated through the second and third minicatheter. Subsequently, physiologic solution was infused through the first minicatheter for washout. Using the same method, the left corpora was treated with CHAPS through the second minicatheter and then extensively washed with Krebs. The tissues were then dissected and placed in the organ chambers. Strips of CHAPS treated and control corporal tissue were contracted with phenylephrine (two µM) in parallel. When the tension had stabilized, cumulative concentrations of acetylcholine (one nM to 100 µM) were added to the organ chamber as a functional test of endothelial integrity. Only the CHAPS treated tissues in whom the maximal relaxation to acetylcholine was 15% or less were considered to be significantly denuded of endothelium. Of the 36 animals studied, removal of endothelium with CHAPS was successful in 53%. The corporal tissue from these animals contracted with phenylephrine but the relaxation to acetylcholine was greatly reduced or completely blocked (figure 1). In 30% of the CHAPS treated corpora, relaxation of the corporal smooth muscle to acetylcholine was similar to control tissues. Based on this observation, the tissues from these animals were rejected from the study and were considered as a failure to remove the endothelium. In 17% of the animals treated with CHAPS, the tissue did not contract with phenylephrine indicating damage to the smooth muscle. These animals were also rejected from the study. Identification of the specific endothelial marker, factor VIII, using immunocytochemical techniques was attempted with light microscopy to confirm histologically the removal of endothelium with CHAPS. Because the rabbit corpus cavernosum endothelium is very thin, we did not succeed in determining the presence or absence of endothelium with this technique. For this reason, transmission electron microscopy was used. Several (approximately 10) sections were examined from different regions of the corporal strips in an attempt to get an overall evaluation of the structure of endothelium in control and CHAPS treated strips. In the CHAPS treated tissue, complete removal or disruption of endothelium was observed in the majority of the corporal lacunar spaces. Nitrnprnsside Hog M)

Acetylcholiile (-log M)

CHAPS

9g

7

I

I

w

5

¥

9

i /// 0 ~

t~ I I ~ t I

V•V

9 I II 7

t

7

1111

,

I

;

b.5 o.o 4.5

•,L 5 min.

9 111 I

I

;

I t

7

11 1, 5 CONTR O L w ~ o I~ , I It //j 5 I I I

t

Phenylephrine {2xJo-6 M)

t

t Phenylephrine (2x10-6 M)

FIG. 1. Tracing represents responses of CHAPS treated and control rabbit corpus cavernosum muscle to acetylcholine and sodium nitroprusside. Contraction was induced by phenylephrine (2 µM). Concentrations of acetylcholine and sodium nitroprusside are shown as negative logarithm of molarity (-log M).

222

AZADZOI AND ASSOCIATES

However, in some lacunar spaces endothelium persisted despite treatment with CHAPS (fig. 2). Cyclic guanosine monophosphate (cGMP) measurement. Rabbit corpus cavernosum strips were submerged in 10 ml. organ baths with Krebs-Ringer's solution containing indomethacin (three µM) and bretylium (10 µM). Methylene blue (O.l mM) was added to the study tissue and parallel untreated strips were utilized as control. The tissues were equilibrated for 30 minutes followed by the addition ofphenylephrine (one µM) to the bath. After another 20 minute equilibration period, tissues were treated with acetylcholine (10 µM) or nitric oxide (0.1 µM) for 15 seconds. Parallel untreated strips were used as controls. Tissues were removed from the bath and submerged into liquid nitrogen. Samples were stored at -80C until analyzed. For cGMP assay, partially thawed tissues were homogenized in one ml. of 6% trichloroacetic acid with a glass-glass tissue grinder on ice. The crude extracts were centrifuged at 3,000 g for 30 minutes at 2C. The supernatants were extracted four times with five volumes of water saturated diethyl ether. The resulting aqueous phase was frozen and lyophilized. Tissue pellets were resuspended in 1 ml diethyl ether and centrifuged at 3,000 g for 15 minutes. After ether removal, the pellets were solubilized at 60C in 0.5 to 1.0 ml of one N NaOH. Samples were diluted to 0.1 N NaOH and protein assay was performed by the method of Lowry. 14 Cyclic GMP levels were measured by a commercially available radioimmunoassay system (Amersham Corp., Arlington Heights, IL). Eicosanoids measurement. Rabbit corporal strips were placed in 24 well plates in two ml. of Dulbecco's modified medium with 0.05% albumin with or without indomethacin (one µM), pH 7.4, 37C, in an air/CO2 tissue culture incubator. Following three 30 minute washout periods the strips were exposed to either one µM acetylcholine, 100 nM VIP, one µM indomethacin or no treatment (time control). Following a 30 minute incubation, the medium was removed and substituted with 10 µM acetylcholine, one µM VIP, one µM indomethacin or no treatment, respectively. After a second 30 minute incubation period, the media were removed, tissues were blotted, dried, weighed, and frozen. Samples were frozen until analyzed. The quantity of eicosanoids released by the corporal tissue into the medium was determined by radioimmunoassay. The level of 6Keto-PGF 1 alpha (stable hydrolytic product of PGI2), thromboxane B2 (stable hydrolytic product of thromboxane A2 ), PGF 2 alpha and PGE 2 were measured. Radioimmunoassays were performed using specific antisera to thromboxane B 2 , PG F 2 alpha, PGE 2, 6-keto-PGF1 alpha, tritiated standards and unlabeled standards. All dilutions were made with physiologic salt solution. Standard curves were obtained with equal volume of physiologic salt solution to that being assayed. Cross reactivity with other measured prostanoids was