SEARCH

SEARCH BY CITATION

Keywords:

  • cyclic nucleotides;
  • cGMP;
  • cGMP-dependent protein kinase;
  • lower urinary tract;
  • nitric oxide;
  • neurotransmission;
  • smooth muscle

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

OBJECTIVES

To characterize the distribution of phosphodiesterase 5 (PDE-5), cGMP and cGMP-dependent protein kinase I (PKG1), and to evaluate the effect of pharmacological inhibition of PDE-5 in isolated preparations of pig and human urethra, as the nitric oxide (NO)/cGMP pathway generates the main inhibitory signals to reduce resistance in the bladder outlet and urethra during emptying of the bladder.

MATERIALS AND METHODS

After obtaining ethics committee approval, urethral specimens were obtained from three female patients during cystectomy, and from young female pigs. The specimens were prepared for immunohistochemical investigations and for functional studies in organ baths. Effects of sildenafil, vardenafil and tadalafil (1 nm to 30 µm) were studied in l-noradrenaline (1 µm)-activated or spontaneously contracted preparations, and on relaxations induced by electrical-field stimulation (EFS). Levels of cGMP were determined by radioimmunoassay.

RESULTS

After stimulation with the NO donor, DETA NONO-ate (1 mm), there was greater cGMP-immunoreactivity (IR) in urethral and vascular smooth muscles. There was a wide distribution of cGMP- and vimentin-positive interstitial cells between pig urethral smooth muscle bundles. There was also cGMP-IR within NO-synthase-IR endothelium. There was PDE-5 IR within the urethral and vascular smooth muscle cells, but also in vascular endothelial cells that expressed cGMP-IR. In pig and human sections, there was strong PKG1-IR in α-actin-IR urethral smooth muscle cells that also contained IR for cGMP. Sildenafil, vardenafil and tadalafil caused mean (sem) concentration-dependent relaxations of the pig urethra which, at 30 µm, were 80 (3)% (11 samples), 81 (5)% (12 samples) and 64 (4)% (10 samples) of the spontaneous tone. The relaxation of L-noradrenaline-contracted female human urethra was 100% in response to 10 µm sildenafil, and 85 (15)% and 47 (13)% for 30 µm of vardenafil and tadalafil, respectively (three samples). Vardenafil or sildenafil (30 µm) doubled cGMP levels in pig specimens. There were no effects on cGMP levels with tadalafil. EFS (1–32 Hz) caused l-NG-nitroarginine-sensitive relaxations of pig urethral muscle that were increased in amplitude and duration by PDE-5 inhibition. At 0.1 µm, sildenafil, vardenafil or tadalafil significantly prolonged the mean (sem) duration of the relaxation at 4 Hz by 55 (19)%, 45 (14)% and 51 (12)%, respectively.

CONCLUSIONS

PDE-5-, cGMP- and PKG1-IR is widely distributed in human and pig urethral tissues. Nerve-induced relaxations of urethral preparations were enhanced at low concentrations of sildenafil, vardenafil and tadalafil, whereas there were direct smooth muscle-relaxant actions of the PDE-5 inhibitors at high concentrations. Inhibition of PDE-5 might be an interesting option to facilitate cGMP-mediated relaxation of the outflow region.


Abbreviations
IHC

immunohistochemistry

IR

immunoreactivity

FITC

fluorescein isothiocyanate

EFS

electrical field stimulation

L-NNA

l-NG-nitroarginine

(e)NO(S)

(endothelial) nitric oxide (synthase)

(s)(p)GC

(soluble) (particulate) guanylate cyclase

PDE

phosphodiesterase

PKG1

cGMP-dependent protein kinase I

SIN-1

3-morpholinosydnoimine

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The outflow region of the lower urinary tract contains many nitric oxide synthase (NOS)-containing nerves [1]. NO activates soluble guanylate cyclase (sGC) to produce cGMP, the main intracellular second messenger that produces relaxation of urethral smooth muscle [1,2]. A role for cAMP in direct smooth muscle regulatory functions of the urethra might be considered less important, as nerve-mediated l-NG-nitroarginine (L-NNA) and ODQ-sensitive relaxant responses in urethral preparations from various species are not accompanied by changes in the tissue levels of cAMP [1]. Cyclic nucleotides have also been proposed to be involved in the regulation of detrusor tone. Thus, adenylate cyclase-coupled muscarinic M2 receptors have been suggested to oppose β-adrenoceptor mediated activation of adenylate cyclase, thereby decreasing intracellular concentrations of cAMP and promoting contraction [1]. NO and cGMP appear to have no immediate smooth muscle-related effects in the detrusor [1].

Downstream from their production, the intracellular levels of cAMP and cGMP are regulated by cyclic nucleotide phosphodiesterase (PDE) enzymes. To date, 11 different families of PDE are known, that differ in their specificity for cAMP and cGMP, cofactor requirements, and kinetic properties [3]. Transcription of genes encoding for various PDEs have been detected in urogenital tissues, including human and porcine detrusor smooth muscle, human prostate, penis, clitoris and vagina [4–8].

Relaxant responses to vinpocetine, an inhibitor of the cAMP-degrading PDE-1, were reported for human and porcine detrusor [4,5]. Beneficial effects of vinpocetine were reported on urgency and urgency incontinence in patients not responding to standard pharmacological therapy [9]. Inhibition of cGMP-degrading PDEs seems to have negligible effects on contractile function in detrusor preparations [4,5]. In the urethra, information on the effects of selectively inhibiting PDE-5 on contractile mechanisms of the urethra is scarce. Zaprinast, a relatively selective PDE inhibitor, increases tissue-levels of cGMP and potentiates relaxant effects during activation of nerves in urethral preparations [10,11]. Sildenafil, vardenafil and tadalafil are considered highly selective for PDE-5 [12–14], and sildenafil has been suggested to improve not only erectile function, but also LUTS in men with erectile dysfunction and BPH [15].

The aim of the present study was therefore to further evaluate the NO/cGMP/PDE-5 system in the female pig and human urethra, focusing on the effects of inhibiting PDE-5 on smooth muscle tone and nerve-evoked relaxations. In addition, the distribution of cGMP, cGMP-dependent protein kinase I (PKG1), and PDE-5 were studied using immunohistochemistry (IHC).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The bladder and urethra from young female pigs were acquired from a slaughterhouse and transported to the laboratory in ice-cold Krebs solution. For functional studies, circular smooth muscle strips (1 × 2 × 6 mm) were dissected ≈ 4 cm below the ureteric orifices [16].

After approval of the local Ethics Committee and informed consent, human urethra was obtained from three women patients (mean age 67 years, sd 4) during cystectomy for bladder cancer. Urethral smooth muscle preparations (1 × 2 × 6 mm) were taken longitudinally from the mid-portion of the urethra. Preparations from the respective patient had similar morphological appearances and had similar functional responses in vitro.

For IHC, urethral specimens were immersion-fixed (4 h) in an ice-cold solution of 4% formaldehyde in PBS at pH 7.4 and further processed for IHC as previously described [17]. The formation of cGMP was stimulated as described by Smet et al.[18], with modifications according to Gillespie et al.[19].

Cryostat sections were cut at 10 µm and incubated for 24 h with a rabbit antibody for endothelial NOS (eNOS, 1 : 1000), a mouse antibody for α-actin (1 : 2000), a goat antibody for vimentin (1 : 200, Sigma Chemicals, St Louis, MO, USA), a goat antibody for PKG1 (1 : 200, Santa Cruz Biotechnology, CA, USA), a rabbit antibody for PDE-5 (1 : 200, FabGennix, Shreveport, CA, USA), or a rabbit antibody for cGMP (1 : 200, Biogenesis Ltd. Poole, UK). A sheep cGMP antiserum (a generous gift from Dr J. de Vente, Maastricht University, the Netherlands) was also used (1 : 1000). For the simultaneous detection of two antigens, sections were incubated overnight with a mixture of two appropriate primary antibodies (see above). After incubation with the primary antisera, the sections were rinsed and incubated with the appropriate species directed Alexa Fluor (Molecular Probes, Eugene, OR, USA) green or red fluorescein isothiocyanate (FITC)-labelled secondary antibody (dilution 1 : 400 and 1 : 1000, respectively) for 90 min, rinsed, and mounted. The sections were examined using a light microscope.

For the experimental in vitro system, preparations were mounted in 5 mL aerated (95% O2 and 5% CO2) tissue baths (37 °C) containing Krebs solution. Mechanical activity was registered with force transducers connected to a polygraph (FT 03C and model 7E, Grass Instruments, Quincy, MA, USA). Two platinum electrodes were placed in parallel to the strips in the tissue baths, and electrical-field stimulation (EFS) was applied using a stimulator (Grass S48) delivering square-wave pulses of 0.3 ms duration at 1–32 Hz; the voltage was supramaximal, the train duration 5 s and the stimulation interval was 2 min. Responses to EFS were studied in the presence of 1 µm phentolamine, scopolamine and propranolol, as previously described [16].

For functional experiments in the pig urethra, preparations were stretched to a tension of 10 mN and left to equilibrate for 45 min. To establish the ‘maximum’ relaxation level, preparations were exposed to Ca2+-free Krebs solution [16]. A stable tension of 8.4 ± 0.7 mN (10 strips) was then established with Ca2+-containing Krebs solution. The effects of vardenafil, tadalafil and sildenafil (1 nm to 30 µm) on spontaneous tone were assessed. When a stable relaxation was obtained in 30 µm of the respective agent, preparations were snap-frozen in liquid nitrogen to determine cAMP and cGMP levels.

The effect of 0.1 µm sildenafil, vardenafil or tadalafil on the amplitude and duration of EFS-induced relaxations was assessed. At this concentration there was < 10% loss of tone with the respective PDE-5 inhibitor. The duration of EFS-induced responses was assessed at 4 and 16 Hz, and calculated at half the relaxant amplitude. In addition, effects of 3-morpholinosydnoimine (SIN-1, 10 nm to 10 µm) or forskolin (1 nm to 1 µm) were studied on the spontaneous urethral tone before and after 0.1 µm sildenafil, vardenafil or tadalafil.

For functional experiments in the human urethra, strips of female human urethra were given a passive tension of 4–5 mN, and left to equilibrate for 40 min. To study viability and contractile capacity, the preparations were exposed to K+ (124 mm) Krebs solution. The effects of the cumulative addition of vardenafil, tadalafil and sildenafil (1 nm to 30 µm) on 3 µm l-noradrenaline-contracted preparations were assessed.

For cGMP and cAMP measurements, the tissue was homogenized, centrifuged and processed as previously described [16]. The pellet was processed for protein content, and the aqueous phase was evaporated and the residue used in 125I-cGMP and 125I-cAMP radioimmunoassay kits (DuPont, Wilmington, DE, USA). The final values of tissue cAMP were corrected for trace amounts of 3H-cAMP.

The following drugs were used: l-noradrenaline, L-NNA, scopolamine, phentolamine, propranolol, SIN-1 (all Sigma Chemicals), DETA NONO-ate (Alexis Corp., Laussen, Switzerland), and sildenafil (Pfizer Ltd, UK) were dissolved in NaCl. Forskolin (Sigma) was dissolved in ethanol. Stock solutions of vardenafil and tadalafil (Abbott Laboratories, Abbot Park, IL, USA) were dissolved in 50% ethanol and DMSO, respectively; subsequent dilutions were made in NaCl. The Krebs solution had the following composition (mM): NaCl 119, KCl 4.6, CaCl2 1.5, MgCl 1.2, NaHCO3 15, NaH2PO4 1.2, glucose 5.5. Ca2+-free Krebs solution was prepared by omitting CaCl2 and adding EGTA (0.1 mm). K+- (124 mm) Krebs solution was prepared by replacing NaCl with equimolar K+.

Relaxations of spontaneous tone are expressed as a percentage of the maximum relaxation level. The relaxant effects in human urethra are expressed as a percentage of the l-noradrenaline-induced tension. The concentration of drug producing half the relaxation of spontaneous or agonist-induced tension was calculated and expressed as the − log IC50.

Values are given as the mean (sem), with Student's two-tailed t-test used for paired or unpaired observations. For multiple comparisons, a one-way anova (Bonferroni) was used, and in all tests P < 0.05 was regarded as indicating significance, with the number of pigs used (or human samples) given in parentheses.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

There was weak cGMP-immunoreactivity (-IR) within smooth muscle of the pig urethra in sections not exposed to DETA NONO-ate. In sections of the pig urethra stimulated with DETA NONO-ate, cGMP-IR increased in smooth muscle cells of the urethra. In addition, there were strong cGMP-IR signals in small and slender cells between bundles of smooth muscle cells (Fig. 1). The former cells had central nuclei and extended dendritic processes that formed interconnecting networks around the smooth muscle bundles and occasionally around arteries. In double-labelled sections, the dendritic cGMP-IR cells also contained vimentin-IR (Fig. 1) but not PDE-5-IR.

image

Figure 1. IHC: IR for cGMP in interstitial cells and smooth muscle cells in a DETA NONO-ate-stimulated female pig urethral section. (a) × 10, FITC immunofluorescence; (b) the same section as in (a) but at × 40; (c) × 40, Alexa Green immunofluorescence; (d) the same section as in (c) but at vimentin-IR in interstitial cells; Alexa Red immunofluorescence, ×40.

Download figure to PowerPoint

In pig and human sections there was strong PKG1-IR in α-actin-IR urethral smooth muscle cells that also contained cGMP-IR (Figs 1 and 2). Whereas there was only weak PDE-5-IR detected in smooth muscle cells of the pig urethra, there was strong PDE-5-IR in smooth muscle cells of the human urethra (Fig. 2).

image

Figure 2. IHC: Human female urethra: (a) IR for cGMP in smooth muscle cells. × 20, Alexa Green immunofluorescence; (b) the same section as in (a) but showing IR for PKG1, Alexa Red immunofluorescence, ×20. (c) IR for PKG1 in smooth muscles of the urethral wall, × 10, Alexa Green immunofluorescence; (d) the same section as in (c) but showing α-actin-IR, Alexa Red immunofluorescence, ×10; (e) IR for PDE-5 in smooth muscle bundles of the urethra, × 10, Alexa Green immunofluorescence; (f) the same section as in (e), but for α-actin-IR, Alexa Red immunofluorescence, ×10.

Download figure to PowerPoint

In sections of the pig urethra not exposed to DETA NONO-ate there was strong cGMP-IR in eNOS-IR endothelial cells of arteries and arterioles in the urethral wall or sub-urothelially (Fig. 3). There was weak cGMP-IR in the smooth muscle wall of these vessels. There was strong PKG1-IR within cGMP-IR smooth muscle bundles of arteries and arterioles in human (Fig. 3) and DETA NONO-ate-stimulated pig urethral sections. In human and pig urethral sections there was PDE-5-IR in smooth muscle cells of the wall of larger arteries and in endothelial cells lining the arterial lumen, whereas there was PDE-5-IR only in the endothelium of small arterioles (Fig. 3). All the human tissues showed a similar distribution of cGMP-, PKG1- and PDE-5-IR.

image

Figure 3. IHC. IR for cGMP in endothelial cells: (a) of a small muscular artery in the urethral wall of (a) the female pig urethra, × 40, Alexa Green immunofluorescence; (b) the same section as in (a) but for IR for eNOS, Alexa Red immunofluorescence, ×40; (c) of the smooth muscles of a small muscular artery of the female human urethral wall, × 40, Alexa Green immunofluorescence; (d) the same section as in (c) but for PKG1-IR, Alexa Red immunofluorescence; (e) of a suburothelial artery of the female pig urethra, × 40, Alexa Green immunofluorescence; and (f) the same section as in (e) but for PDE-5; (g) IR for PDE-5 in a suburothelial arteriole. U denotes urothelium, × 40, Alexa Green fluorescence.

Download figure to PowerPoint

Sildenafil, vardenafil and tadalafil (1 nm to 30 µm) caused concentration-dependent relaxations of urethral tone (Fig. 4). The relaxations evoked by 30 µm of sildenafil, vardenafil, or tadalafil reached 80 (3)% (11 samples), 81 (5)% (12), and 64 (4)% (10; P < 0.05 vs sildenafil and vardenafil). The – log IC50-values were 5.19 (0.06), 5.63 (0.17) and 5.17 (0.13) for sildenafil, vardenafil and tadalafil, respectively.

image

Figure 4. (a) Concentration-response curves to 10 nm to 30 µm sildenafil (11 samples; circles), vardenafil (12; squares) and tadalafil (10; triangles). Values are the mean (sem); tracings show the relaxant effects of (b) sildenafil (10 nm to 30 µm), (c) vardenafil (10 nm to 10 µm) and (d) tadalafil (10 nm to 30 µm) in spontaneously contracted smooth muscle strips from the female pig urethra. Lines indicate the maximum relaxation level (baseline).

Download figure to PowerPoint

SIN-1 (1 nm to 10 µm) concentration-dependently relaxed the pig urethra to a maximum of 92 (5)% (six samples; Fig. 5) at 10 µm; the – log IC50 for SIN-1 was 5.83 (0.10) before, and 6.41 (0.08) after treatment with sildenafil (six; P < 0.001). Vardenafil changed the – log IC50 for SIN-1 from 6.24 (0.11) to 6.91 (0.21) (six; P < 0.01). Tadalafil (0.1 µm) changed the − log IC50 for SIN-1 from 5.93 (0.11) to 6.60 (0.10) (six; P < 0.01). Forskolin caused a maximum relaxation of 100% (six samples) at 1 µm; sildenafil, vardenafil or tadalafil did not affect the relaxant effect of forskolin.

image

Figure 5. Relaxant effects of the NO-donor SIN-1 (10 nm to 10 µm) in the female pig urethra in the absence (red circles) and presence (green circles) of sildenafil 0.1 µm. Values are the mean (sem).

Download figure to PowerPoint

EFS relaxed the pig urethral smooth muscle tone (Fig. 6); treatment with L-NNA (0.1 mm) abolished the relaxation at low frequencies and reduced them at higher frequencies. In the presence of sildenafil, the amplitude of the maximum relaxation evoked by EFS (32 Hz) was 45 (6)%, vs 38 (4)% before treatment (Fig. 6a, 14 samples, P < 0.05). The maximum relaxation was 35 (3)% before treatment with vardenafil and 39 (5)% afterward (Fig. 6b, 11 samples); the corresponding values for tadalafil were 30 (2)% before treatment and 36 (4)% afterward (Fig. 6c, nine samples; P < 0.05).

image

Figure 6. Enhancing effects on the relaxations evoked by EFS (1–32 Hz) by 0.1 µm of(a) sildenafil (14 samples), (b) vardenafil (11) and(c) tadalafil (nine). Red symbols are controls, Green symbols are the PDE-5 inhibitor. Values are the mean (sem). *P < 0.05.(d) Tracing showing the relaxant effects of EFS (4 and 16 Hz) in the female pig urethra. The addition of tadalafil (0.1 µm) increased the amplitude and duration of the relaxation at 4 and 16 Hz, and the relaxation at 4 Hz was abolished by 0.1 mm l-NNA.

Download figure to PowerPoint

Treatment with sildenafil, vardenafil and tadalafil increased the duration of the relaxations evoked by EFS (Table 1, Fig. 6). Sildenafil significantly prolonged the duration of the EFS-induced relaxation at 4 Hz by 55 (19)%, and at 16 Hz by 56 (16)% (14 samples), vardenafil did so by 45 (14)% and 60 (16)%, respectively (10 samples) and tadalafil by 51 (12)% (P < 0.05) and 26 (12)%, respectively (nine samples, Fig. 6d).

Table 1.  Effects of 0.1 µm sildenafil (14 samples), vardenafil (10) and tadalafil (nine) on the duration of the electrically evoked relaxation at 4 and 16 Hz of stimulation. Values are calculated at half the maximal amplitude of the relaxation, and are the mean (sem)
PDE-5 inhibitorStimulation frequency, HzDuration, s
before treatmentafter treatment
  • *

    P < 0.05, control vs treatment with PDE inhibitor.

Sildenafil 414.4 (1.4)19.3 (1.8)*
1618.9 (1.4)26.4 (2.0)*
Vardenafil 415.0 (1.8)20.2 (1.4)*
1619.0 (1.3)28.2 (2.4)*
Tadalafil 415.4 (2.0)21.7 (2.4)*
1625.6 (1.4)30.0 (1.8)

In human urethra, sildenafil, vardenafil and tadalafil (1 nm to 30 µm) concentration dependently relaxed l-noradrenaline-contracted urethral preparations (Fig. 7). The l-noradrenaline-induced contraction was abolished in strips exposed to 10 µm sildenafil (three). At the same concentration, the relaxation to vardenafil or tadalafil was 85 (15)% and 47 (13)% (both three), respectively. The maximum relaxation was 94 (6)% and 85 (10)% in strips exposed to 30 µm vardenafil or tadalafil. The − log IC50 was 5.82 (0.08) for sildenafil, 5.82 (0.40) for vardenafil and 5.14 (0.56) for tadalafil.

image

Figure 7. (a) Concentration-response curves to 10 nm to 30 µm sildenafil (three samples; circles), vardenafil (three; squares) and tadalafil (three; triangles) in human urethra. Values are the mean (sem). (b) a tracing showing the relaxant effects of vardenafil (1 nm to 10 µm) in a human smooth muscle strip pre-contracted with 3 µm l-noradrenaline.

Download figure to PowerPoint

Vardenafil (30 µm) increased cGMP levels in pig urethral strips from 8.5 (1.6) to 16.4 (2.5) pmol/mg protein (six samples, P < 0.05; Fig. 8). Sildenafil 30 µm increased cGMP levels to 14.2 (1.2) pmol/mg protein (five, P < 0.05) but there were no effects on cGMP levels after treatment with 30 µm tadalafil, at 8.0 (1.1) pmol/mg (six; Fig. 8). In control experiments, treatment with DMSO had no effect on cGMP levels, at 6.8 (2.0) pmol/mg (six). Sildenafil, vardenafil or tadalafil did not affect the levels of cAMP.

image

Figure 8. The cGMP content in pig urethral tissue at resting tension (control, six samples) and in the presence of 30 µm sildenafil (five), vardenafil (six) and tadalafil (six). *P < 0.05 vs control.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The GCs catalyse the conversion of GTP to cGMP, which is an important intracellular signalling molecule involved in the regulation of, e.g. motility and smooth muscle contraction, fluid and electrolyte homeostasis, and phototransduction [20]. The cellular response to increased levels of cGMP depend upon the distribution of target proteins, the presence of parallel and co-active transduction pathways, and regulatory systems of the cGMP-initiated signal [3,20]. In the female human and pig urethra, cGMP-IR was detected in smooth muscle cells that also expressed IR for PKG1 and PDE-5. In an environment rich in NO and during simultaneous inhibition of PDEs, the intensity of cGMP-IR was substantially increased in pig urethral smooth muscles. In these specimens, there were spindle-shaped cGMP-IR and vimentin-positive cells that formed a network around and between the smooth muscle bundles. Similar cGMP-IR cells were reported in guinea-pig [17–19] and human [18] urinary bladder and urethra but have, to our knowledge, not previously been reported for the pig urethra. The function of the urethral interstitial cells has not been established, but they have been suggested to be pacemaker cells involved in the regulation of tone [21].

Corresponding to the morphological findings, sildenafil, vardenafil or tadalafil relaxed female pig and human urethra without the simultaneous presence of a cGMP-producing stimulus. This is in accordance with findings that zaprinast by itself caused a reduction of noradrenaline-induced tone of the female rabbit urethra, and suggests basal activity of the cGMP pathway within structures in the urethra [11]. Also, under the current experimental conditions, pig or human urethral specimens not stimulated by cGMP expressed IR for cGMP in smooth muscle cells and in vascular endothelium. In support of a role for NO in maintaining basal activity in the cGMP signal of the urethra, the NO-synthesis inhibitor L-NNA was previously shown to decrease the tissue levels of cGMP of preparations from the female pig urethra [16]. However, inhibiting the synthesis of NO failed to affect the spontaneous tone of preparations of the female pig urethra, and only ≈ 15% of preparations from the outlet region of the same species responded to L-NNA with additional increases in agonist-induced tension [11,16]. Possibly additional upstream activators of the sGCs have effects on the basal levels of cGMP in the urethra. Carbon monoxide may be such one candidate that, similar to NO, might activate the sGC [17,22,23]. The CO-producing enzyme, heme oxygenase was located in neuronal structures of the pig urethra, and CO evokes relaxant effects in this preparation that are associated with increases in cGMP, but not cAMP [22]. cGMP might also be produced by the particulate GCs (pGCs), located in the urethra and bladder of guinea pigs [24]. Activation of the pGC by atrial natriuretic peptide produces relaxation of urethral preparations from the guinea-pig, with concomitant increases in the tissue levels of cGMP [24].

The mechanism by which sildenafil, vardenafil and tadalafil produce relaxations of the urethra is presumably by inhibiting PDE-5 and subsequent accumulation of cGMP. This is supported by the findings that vardenafil and sildenafil almost doubled the cGMP tissue levels in the pig urethra. However, that sildenafil or vardenafil per se caused elevations of cGMP in urethral tissue is in contrast to findings in rabbit or human corpus cavernosum, where sildenafil (0.01–1 µm) alone did not increase the tissue content of cGMP [25,26]. When examining the tissue levels of cGMP in the present study, the concentration (30 µm) of the respective PDE-5 inhibitor that produced the largest relaxant responses was used. This might account for better drug availability at the level of PDE-5, but might still not be expected to yield increases in cGMP unless there is basal activity of the GCs.

Similar to the present study, the sGC-activator YC-1 alone caused a similar change in the ratio of the urethral tissue levels of cGMP vs untreated control preparations but without affecting the spontaneous or agonist-induced tension [23]. If the levels of cGMP reflect the amplitude of relaxations evoked by each of the PDE inhibitors in the present investigation, tadalafil would not be expected to increase levels of cyclic nucleotides to those with sildenafil and vardenafil, which at any concentration produced a 10–35% larger relaxant response than did tadalafil. Surprisingly, tadalafil did not affect the levels of cGMP. This cannot be accounted for by actions of the vehicle, as DMSO alone had no effect on the tension of the preparations or on the tissue levels of the cyclic nucleotides. PDE-5 inhibitors might produce effects partly through other than cGMP-dependent pathways. In the human pregnant myometrium, sildenafil in millimolar concentrations was suggested to act on tetraethyl ammonium-resistant potassium channels [27]. Direct effects of sildenafil at prejunctional large-conductance Ca2+-activated K+ channels were reported in human vas deferens, and sildenafil was also suggested to open ATP-sensitive potassium channels in endothelial cells [28,29]. However, the structural differences between sildenafil, vardenafil and tadalafil might not allow for such assumptions in the pharmacological properties of this class of agents [13,30]. Sildenafil and vardenafil are of similar size and structure, and only differ in the core-ring system and a methyl-/ethyl group at the piperazine ring, whereas tadalafil is developed from a structurally different series of PDE-5 inhibitors [13,31]. The greater potency of vardenafil than sildenafil for inhibiting PDE-5 is proposed mainly to be due to the differences in the ring systems between the compounds [31]. Tadalafil was reported to have a fast dissociation rate from PDE-5 [30] that might be related to the difference in the pharmacodynamic properties of this PDE-5 inhibitor.

Exogenously applied NO-donors cause relaxations of urethral preparations with simultaneous increases in the tissue contents of cGMP [2]. In the female pig or rabbit urethra, sodium nitroprusside, SIN-1, NaNO2, or 8-Br-cGMP effectively relax spontaneously active or agonist-contracted preparations [11,32]. Irrespective of the agent used, inhibition of PDE-5 during cumulative addition of SIN-1 in preparations of the female pig urethra significantly enhanced relaxant responses to the NO-donor. Without affecting the maximum relaxant effect by SIN-1, there were significant leftward shifts in the concentration-response curves to the agent in the presence of sildenafil, vardenafil or tadalafil. Potentiating effects of 50–100% of the relaxant responses to SIN-1 were recorded at 0.1 µm to 3 µm of SIN-1.

It is generally accepted that nerve-released NO is important for the relaxation of urethral smooth muscle in several species, and that this relaxation is accompanied by larger amounts of cGMP [1]. Interventions to increase the endogenous production of cGMP could therefore have the potential to affect nerve-induced relaxations. At various frequencies, 0.1 µm of sildenafil or tadalafil increased the amplitude of nerve-mediated relaxations by 20–30%. This is in accordance with previous findings that zaprinast, at a low submaximal frequency, increased the amplitude of relaxations during EFS [10]. In the present study, pre-treatment with vardenafil showed tendencies to increase the amplitude of nerve-induced relaxations in the lower range of the frequencies tested. Favouring the view that cGMP-mediated mechanisms might be pharmacologically targeted to enhance nerve-induced relaxations of the female pig urethra, YC-1 (10 µm) caused a > 150% increase in the amplitudes of EFS-induced relaxant responses in both spontaneously contracted preparations and in agonist-activated strips [23]. In contrast, Persson and Andersson [11] reported that zaprinast (10 µm) failed to increase the amplitude of nerve-mediated relaxations of the rabbit urethra despite a greater accumulation of cGMP in the tissue. Supported by earlier reports on nerve-induced relaxation of the rabbit urethra it was proposed that at the frequency used (12 Hz), a maximum relaxant response was already achieved [11,32].

In the rabbit urethra and the sheep bladder outlet regions, zaprinast was reported to increase the duration of the relaxant responses to EFS [11,32,33]. In the current investigation, sildenafil, vardenafil or tadalafil increased the duration of EFS-induced relaxations by 30–60%. Although these findings cannot be translated to an in vivo situation, it is possible that pharmacological inhibition of PDE-5, mainly by effects in the urethra, can enhance cGMP-mediated relaxations and thereby improve the voiding mechanism with no effects on continence. If given orally, the plasma peak concentration of sildenafil is ≈ 1044 ng/mL at the therapeutic dosage [34], and the expected concentration of the drug in the urethra would be lower than the micromolar concentrations needed to cause relaxation in the present study. Intravenous administration of zaprinast was reported to have facilitatory effects on reflex-evoked urethral relaxations in vivo in the anaesthetized female rat [35]. A non-randomized study with no placebo reported that on-demand treatment with sildenafil to men with erectile dysfunction, LUTS and BPH improved their symptom scores [15].

Altered functions of the NO/cGMP system of the urethra were reported in various experimental models. Mice with targeted depletion of the genes that code for neuronal NOS have dysfunctional bladder outlets [36]. The urethra from mice that lack PKG1 do not show NO/cGMP-mediated relaxations and showed voiding disturbances during awake cystometry [37]. In urethral preparations from diabetic rabbits, reduced NOS-binding sites, reduced cGMP-generating capacity, and diminished nerve-mediated relaxations were reported [38,39]. In urethral specimens from aged guinea-pigs there were 45–64% reductions in the activities of the pGC or sGC, and a concomitant increase in the effective dose for half maximal activity for sodium nitroprusside to stimulate the sGC [24]. Impaired neurogenic-, NO- and cGMP-mediated relaxations of the rabbit urethra were also described under conditions of ischaemia [40]. No information is available on whether altered functions of the NO/cGMP pathway in the human urethra are correlated with the development of LUTS, but diabetic vesico-urethral dysfunction was reported to correlate with abnormal nerve conduction velocities [41].

A role for the NO/cGMP pathway in regulating blood flow to the urethra and bladder can also be considered. In arteries from the pig LUT, acetylcholine was previously shown to effectively produce L-NNA-sensitive relaxant responses [42]. In the present study, cGMP-IR was located in the endothelium and smooth muscle of suburothelial arteries, and in arteries of the wall of the urethra from female pigs and humans. cGMP-IR was co-located with eNOS in vascular endothelium. Supporting a role for PDE-5 in the control of cGMP-mediated vascular effects of the urethra, the enzyme was located by IHC in smooth muscle and endothelium of arteries of various sizes.

In summary, we report the presence of cGMP, PKG1 and PDE-5 in smooth muscle cells of the female human and pig urethra. Furthermore, inhibition of PDE-5 with sildenafil, vardenafil or tadalafil amplified endogenous and exogenous NO/cGMP-mediated relaxant effects in the urethra. We propose that selective inhibition of PDE-5 might have facilitatory effects on nerve-mediated relaxation of the urethra and thereby might improve the voiding mechanism.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

We would like to thank Dr Malcolm Carringer, Department of Urology, Örebro Hospital for the generous supply of human tissue. This work was supported by the Swedish Medical Research Council (grant no 6837) and the Medical Faculty, Lund University.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES