Effect of cyclooxygenase inhibitors on the micturition reflex in rats: correlation with inhibition of cyclooxygenase isozymes

Authors


Patrizia Angelico, Pharmaceutical R&D Division, Recordati SpA, Via M. Civitali 1–20148, Milano, Italy. e-mail: angelico.p@recordati.it

Abstract

OBJECTIVE

To investigate the role of cyclooxygenase (COX) isozymes (COX-1 and -2) in the regulation of bladder volume capacity (BVC) in several rat urodynamic models, using a selection of nonsteroidal anti-inflammatory drugs (NSAIDs), some selective for COX-2, correlating the potency of the tested compounds in the urodynamic models and their in vitro potency as inhibitors of COX isozymes, to verify the relative importance of the different isozymes.

MATERIALS AND METHODS

The effects of an i.v. administration of several nonselective and selective COX-2 inhibitors (indomethacin, meloxicam, naproxen, aspirin, paracetamol, flurbiprofen, nimesulide, NS-398, celecoxib, rofecoxib and L 745337) on bladder filling and voiding were evaluated in conscious and anaesthetized rats by cystometry. The cystometry was done in conscious rats 1 day after catheter implantation, by filling the bladder with dilute acetic acid (0.2%) or saline, and again with saline 5 days after catheterization. Effects on isovolumic bladder contractions in anaesthetized rats were also evaluated.

RESULTS

All the NSAIDs tested dose-dependently increased BVC; their potency at increasing BVC during infusion of the bladder with acetic acid was similar to that evaluated with saline on cystometry 1 day after catheterization. When a nonselective (naproxen) and a selective (nimesulide) COX-2 inhibitor were tested in rats with bladders infused with saline 5 days after catheterization, their effects on BVC were significantly lower than those evaluated at 1 day. All tested compounds dose-dependently inhibited isovolumic bladder contractions in anaesthetized rats. There was a good correlation between the potency in inhibiting the isovolumic bladder contractions in anaesthetized rats and in increasing BVC during cystometry in conscious rats with the bladder infused with acetic acid. The potency of the compounds in the cystometry model with bladders infused with acetic and in the isovolumic bladder voiding contractions correlated well with COX-2 inhibition, but not COX-1.

CONCLUSIONS

Both nonselective and COX-2 selective inhibitors are more active in inhibiting the micturition reflex in rats with bladder overactivity caused by bladder irritation than in normal rats. The potency of the anti-inflammatory compounds in inhibiting bladder overactivity induced by chemical or surgical irritation, and their activity in a cystometrographic model practically independent of bladder irritation (isovolumic bladder contractions in anaesthetized rats), was related to the potency as inhibitors of COX-2 isozyme. This suggests that the involvement of prostaglandins in the micturition reflex in rats is mainly mediated by this isozyme.

Abbreviations
COX

cyclooxygenase

PG

prostaglandin

BVC

bladder (volume) capacity

MP

micturition pressure

DT

disappearance time of contractions

EDX

extrapolated dose inducing an X% increase of BVC.

INTRODUCTION

Available evidence indicates that arachidonic acid metabolites produced along the cyclooxygenase (COX) pathway are involved in the physiological regulation of micturition during reflex activation of the urinary bladder. Prostaglandins (PG) E1, PGE2, PGD2, PGF and thromboxane A2 induce contractile activity of animal and human bladder muscle in vitro [1–3]. A role for PGE2 and PGF in modulating cholinergic and purinergic contractions generated by electrical stimulation was shown in rabbit isolated urinary bladder [4]. Furthermore, endogenous PGs are produced locally after distension of the bladder wall and modulate the afferent branch of the reflex micturition by lowering the threshold for eliciting voiding contractions, serving as a link between detrusor muscle stretch produced by bladder filling and activation of capsaicin-sensitive afferents [1,5–7].

Accordingly, human detrusor muscle in vitro relaxes in response to indomethacin [8], and systemic administration of NSAIDs (e.g. indomethacin, ketoprofen, piroxicam and acetylsalicylic acid) inhibit volume-induced bladder voiding contractions in anaesthetized rats [1], with potencies directly proportional to their anti-inflammatory effectiveness which, in turn, is thought to depend on their ability to inhibit the production of PGs [9]. Furthermore, SC-19220, a competitive PG receptor antagonist, increased the bladder volume capacity (BVC) of urethane-anaesthetized rats [6], and indomethacin and flurbiprofen showed favourable effects in double-blind controlled studies in patients with detrusor instability [10], although they did not completely abolish detrusor overactivity and had a high incidence of side-effects.

COX is the pivotal enzyme in PG biosynthesis, and at least two COX isozymes have been characterized extensively. The constitutive COX-1 is thought to be responsible for the physiological functions of PGs, whereas the inducible COX-2 is involved in inflammation [11]. COX-1 and COX-2 isozymes are both expressed in the urinary bladder of adult rodents [12] and, whilst the levels of COX-1 are not altered by physiological or pathophysiological variations, COX-2 is induced during inflammation or bladder obstruction [13]. Lecci et al. [14] studied the effect of dexketoprofen (a nonselective COX inhibitor) and NS-398 (a selective COX-2 inhibitor) on normal and inflamed bladders, and suggested that COX-1 may be involved in modulating the threshold for activating the micturition reflex in normal rats, and that inhibition of COX-2 prevents the urodynamic changes associated with bladder inflammation.

The aim of the present study was to test this suggestion. The activity of several nonselective COX inhibitors and selective COX-2 inhibitors was studied in conscious and anaesthetized rats using different cystometric procedures involving both normal and irritated bladders. Cystometry was assessed in rats 1 day after intravesical catheter implantation infusing into the bladder saline or dilute acetic acid, as models representative of different degrees of bladder irritation. Cystometry was also used in conscious rats 5 days after surgery, and as in anaesthetized rats, immediately after intraurethral catheterization, as models representative of no bladder irritation.

MATERIALS AND METHODS

Male and female Sprague Dawley rats (Crl: CD°(SD)BR) of 200–400 g body weight (Charles River Italia) were housed with free access to food and water and maintained on a forced light-dark cycle at 22–24 °C, except during experiments. The rats were handled according to internationally accepted principles for the care of laboratory animals (EEC Council Directive 86/609, O.J. no L358, 18/12/86).

Cystometric studies in conscious rats were conducted as reported previously [15,16], with some modifications. Male rats, anaesthetized with equithensin solution (3 mL/kg) i.p., were placed supine and a ≈ 10 mm midline incision was made in the shaved and cleaned abdominal wall. The urinary bladder was gently freed from adhering tissues, emptied and then cannulated, via an incision at the dome, with a polyethylene cannula (Portex, internal diameter 0.58 mm, outside diameter 0.96 mm), which was permanently sutured with silk thread. For i.v. injection, another polyethylene tubing (Portex, same size) filled with physiological heparinized (40 IU/mL) saline was inserted into the jugular vein. The cannulae were exteriorized through a subcutaneous tunnel in the retroscapular area, where they were connected with a plastic adapter to avoid the risk of removal by the rat. On the day of the experiment, the rats were placed in Bollman's cages; after a stabilization period of 20 min, the free tip of the cannula was connected by a T-shaped tube to a pressure transducer (P23 XL, Statham, Gould, Cleveland, OH, USA) and to a peristaltic pump (Minipuls 2, Gilson, Middleton, WI, USA) for continuous infusion of warm saline solution (37 °C) into the bladder, at a constant 0.1 mL/min. Two variables from the cystometrogram were recorded on a polygraph or a computerized data acquisition system, i.e. the micturition pressure (MP), defined as the maximum intravesical pressure induced by the contraction of detrusor during micturition, and the BVC, defined as the bladder volume capacity (amount of fluid infused into the bladder) when detrusor contraction was followed by micturition. The micturition volume was measured only in some experiments by isometric transducers positioned under the Bollman's cages.

In rats 1 day after catheterization the basal BVC and MP were evaluated as the mean values on cystometrograms in the 30-min of saline infusion (basal I) before switching from saline to dilute (0.2%) acetic acid or continuing with saline infusion. Cystometrograms were continuously recorded during these infusions for 1 h. The BVC and MP on the last two cystometrograms of this period were averaged to obtain the BVC and MP before treatment (basal II). Then, the rats were injected i.v. with the test compound (or its vehicle) during continuous infusion of the bladder with saline or irritant, and changes in BVC and MP were evaluated for 1 h. The mean BVC and MP were evaluated considering all the cystometrograms recorded during the 1-h of observation.

In rats 5 days after catheterization, the basal BVC and MP were evaluated as the mean of cystometrograms recorded in the 30-min of saline infusion (basal I). Then the rats were injected i.v. with the test compound (or its vehicle) during continuous infusion of the bladder with saline, and changes in BVC and MP evaluated for 1 h. The mean BVC and MP were evaluated considering all the cystometrograms recorded during the 1-h period. Drugs and vehicles were administered in 1 mL/kg.

The effects of compounds on the isovolumic bladder contractions in anaesthetized rats after i.v. or i.c.v. administration were tested as previously reported [15], with minor modifications. Briefly, female rats were anaesthetized with urethane 1.25 g/kg (5 mL/kg, s.c.), and the bladder was catheterized via the urethra using polyethylene tubing filled with physiological saline. The catheter was tied in place with a ligature around the external urethral orifice and the intravesical pressure was measured by a pressure transducer (P23 XL) and displayed continuously on a chart recorder. The bladder was filled through the recording catheter by incremental volumes of warmed (37 °C) saline until spontaneous contractions occurred (usually 0.8–1.5 mL). The activity of the compounds was assessed after i.v. administration through a polyethylene cannula inserted into the jugular vein, or after i.c.v. administration through a stainless-steel guide cannula implanted into the right lateral cerebral ventricle, in at least 4–6 rats per group at each dose. As the compounds had a relatively rapid effect leading to complete cessation of bladder contractions, the activity was conveniently estimated by the duration of bladder quiescence (disappearance time of contractions, DT). The effects on the amplitude of contraction were also estimated comparing them (when contractions re-started) with those previously recorded for 15 min after i.v. administration in the same rats with vehicle alone.

The effect of indomethacin, meloxicam, naproxen, aspirin, paracetamol, flurbiprofen, nimesulide, NS-398 and celecoxib on cystometrographic variables was evaluated in conscious rats with bladders infused with acetic acid or saline. For these experiments, rats were assessed 1 day after catheterization. In general, three scaled i.v. doses of each compound were administered. In some pilot experiments, each dose was tested with a matched control group injected with vehicle and the experiment conducted over >1 day (e.g. indomethacin, meloxicam, naproxen). In other experiments, control rats and rats injected with different doses of the test compound were studied on the same day, and then a single group of control rats was used (e.g. aspirin, paracetamol, flurbiprofen, nimesulide, NS-398, celecoxib). Naproxen and nimesulide, both at 3 mg/kg i.v., were also tested in rats after 5 days of catheterization, and with bladders infused with saline.

In anaesthetized rats, the effect of the compounds studied in conscious rats, and two additional compounds, rofecoxib and L745337, was evaluated after i.v. administration on isovolumic bladder contractions induced by filling the bladder with saline at a constant suprathreshold volume. Indomethacin and celecoxib were also evaluated after i.c.v. administration.

Indomethacin meglumine salt (Liometacen®, vials containing 50 mg/mL solution) was from Chiesi, Parma Italy. Aspirin lisine salt (Flectadol®, vials containing 500 mg/each) was from Sanofi Winthrop, Milan Italy. Flurbiprofen, naproxen, meloxicam, rofecoxib, L745337 and celecoxib were synthesized in the Chemical Department of Recordati SpA, Milan Italy. Nimesulide was from Vectorpharma, Trieste, Italy, and NS 398 from RBI, Natick USA. Pentobarbital (sodium salt) was from Siegfrid Zofingen, chloral hydrate from Merck (Germany) and urethane from Aldrich (UK). Indomethacin vials and naproxen were diluted in saline, used as vehicle in the control groups. Aspirin vials were dissolved in distilled water. All the other compounds were dissolved in 4% dimethylformamide and 8% Tween 80 (v/v, final concentration) in distilled water. Urethane was dissolved in distilled water. The equithensin solution for anaesthesia had the following composition (g): pentobarbital 1.215, chloral hydrate 5.312, magnesium sulphate 2.657, ethanol 12.5 mL, propylene glycol 49.5 mL, distilled water to 125 mL of final volume.

All data are expressed as the mean (sem). In the cystometry models in conscious rats, the statistical significance of the difference of BVC and MP recorded during basal I, basal II and during the hour after treatment was evaluated by two-way anova (within groups) and Dunnett's t-test. For each rat, values of basal I, basal II and after treatment were calculated as described above. The statistical significance of the difference between the changes of BVC in the group(s) injected with vehicle and those injected with the different doses of each tested compound was evaluated by one-way anova and Dunnett's t-test on the difference (Δ) in values (for each rat, the mean value in the 1-h period after treatment minus that at basal II). The extrapolated dose inducing a 50% or 30% increase of BVC (ED50 for acetic acid infusion and ED30 for saline infusion of the bladder) were evaluated by linear regression analysis. The percentage increase of BVC after administering the different doses of tested compounds was evaluated from the mean value recorded during the 1-h period after administration vs the basal II value, and subtracted by the corresponding percentage change observed in the matched control group. In anaesthetized rats, the extrapolated dose inducing 10 min of bladder quiescence (DT10) values were calculated by linear regression analysis.

RESULTS

After a control period of 30 min with saline infusion, the infusion was switched to 0.2% acetic acid. After 1 h of bladder infusion with the irritant, the BVC in all groups of rats was markedly and significantly reduced (by 45–60%; P < 0.01), indicating bladder hyperactivity and an increase in afferent nerve firing (Table 1). MP (data not shown) was less affected and was not significantly changed by the irritant. In rats injected with vehicle, continuous infusion of the irritant during the second hour induced a further decrease of BVC (generally 10–30%), although this further reduction was not statistically significant.

Table 1.  Acetic acid and saline experiments 1 day after surgery; the effect of i.v. administration of several NSAIDs on BVC (mL) in conscious rats. Values are the mean (sem) recorded for 30 min before starting acetic acid or saline infusion (basal I); at the end of the first hour of acid infusion (basal II); and the mean value in the 1-h period after administration of vehicle or after different doses of the test compounds
Treatments, (dose, mg/kg)Acetic acidSaline
NBasal IBasal II1AfterNBasal IBasal IIAfter
  • 1

    The differences between basal I and basal II in the acetic acid experiments were all significant at P < 0.01.

  • *

    P < 0.05 and

  • P < 0.01 vs basal II.

  • P < 0.05 and

  • P < 0.01 vs vehicle (evaluated on

  • Δ

    values 0–60 min).

  • §

    P < 0.05 vs basal I.

Vehicle 60.56 (0.07)0.30 (0.04)0.23 (0.04)60.58 (0.11)0.65 (0.13)0.66 (0.13)
Indomethacin (0.3) 60.66 (0.12)0.26 (0.09)0.28 (0.08)70.71 (0.13)0.63 (0.11)0.77 (0.11)
Vehicle 70.49 (0.04)0.27 (0.04)0.21 (0.03)70.42 (0.05)0.52 (0.06)0.50 (0.07)
Indomethacin (1.0) 60.67 (0.13)0.29 (0.06)0.32 (0.06)60.39 (0.04)0.46 (0.04)0.60 (0.05)
Vehicle 80.47 (0.05)0.26 (0.04)0.23 (0.03)80.78 (0.09)0.88 (0.11)0.90 (0.14)
Indomethacin (3.0) 80.57 (0.08)0.29 (0.05)0.45 (0.04)*80.69 (0.11)0.69 (0.11)0.97 (0.13)
Vehicle 60.74 (0.10)0.24 (0.06)0.16 (0.03)80.78 (0.11)0.92 (0.15)0.82 (0.11)
Meloxicam (0.3) 60.84 (0.09)0.28 (0.04)0.29 (0.04)80.70 (0.10)0.73 (0.14)0.80 (0.11)
Vehicle 80.70 (0.07)0.32 (0.07)0.23 (0.04)70.91 (0.13)0.94 (0.14)0.89 (0.11)
Meloxicam (1.0) 80.70 (0.09)0.24 (0.08)0.29 (0.05)70.98 (0.09)1.15 (0.10)1.08 (0.09)
Vehicle 70.87 (0.08)0.35 (0.07)0.18 (0.03)70.79 (0.10)0.85 (0.13)0.82 (0.11)
Meloxicam (3.0) 70.91 (0.13)0.36 (0.07)0.58 (0.08)80.73 (0.07)0.77 (0.06)0.93 (0.07)
Vehicle 60.66 (0.08)0.37 (0.08)0.31 (0.07)70.67 (0.12)0.71 (0.15)0.75 (0.18)
Naproxen (0.3) 60.68 (0.08)0.34 (0.05)0.37 (0.03)70.60 (0.08)0.60 (0.09)0.78 (0.09)
Naproxen (1.0) 60.61 (0.11)0.25 (0.05)0.35 (0.05)80.59 (0.06)0.61 (0.07)0.87 (0.08)
Vehicle 80.93 (0.15)0.42 (0.08)0.26 (0.04)80.72 (0.07)0.83 (0.09)0.80 (0.07)
Naproxen (3.0) 80.76 (0.09)0.32 (0.08)0.49 (0.07)*80.59 (0.06)0.66 (0.09)1.09 (0.15)
Vehicle 70.51 (0.09)0.23 (0.04)0.17 (0.02)70.73 (0.11)0.74 (0.11)0.70 (0.08)
Aspirin (10) 50.80 (0.14)0.37 (0.11)0.35 (0.07)70.66 (0.09)0.74 (0.14)0.81 (0.12)
Aspirin (30) 80.65 (0.09)0.27 (0.05)0.39 (0.07)80.76 (0.09)0.84 (0.12)0.93 (0.15)
Aspirin (100) 80.69 (0.08)0.28 (0.06)0.39 (0.06)60.79 (0.14)0.84 (0.16)0.72 (0.13)
Vehicle 90.79 (0.08)0.34 (0.05)0.31 (0.05)80.69 (0.10)0.68 (0.13)0.66 (0.15)
Paracetamol (10) 90.77 (0.10)0.36 (0.08)0.38 (0.08)80.69 (0.09)0.72 (0.10)0.82 (0.09)
Paracetamol (30) 80.76 (0.12)0.31 (0.08)0.37 (0.08)90.69 (0.11)0.71 (0.12)0.87 (0.12)
Paracetamol (100) 80.78 (0.08)0.29 (0.06)0.48 (0.06)80.70 (0.05)0.64 (0.10)0.89 (0.07)
Vehicle 80.58 (0.09)0.27 (0.04)0.22 (0.03)70.60 (0.12)0.64 (0.12)0.65 (0.12)
Flurbiprofen (0.1) 70.57 (0.10)0.27 (0.06)0.36 (0.07)80.52 (0.06)0.57 (0.06)0.69 (0.07)
Flurbiprofen (0.3) 80.76 (0.11)0.32 (0.05)0.46 (0.04)80.79 (0.10)0.96 (0.13)§1.24 (0.11)
Flurbiprofen (1.0) 80.58 (0.10)0.21 (0.03)0.40 (0.05)*90.53 (0.04)0.61 (0.10)0.94 (0.11)
Vehicle 80.59 (0.08)0.27 (0.05)0.25 (0.04)60.69 (0.08)0.72 (0.13)0.61 (0.10)
Nimesulide (0.3) 70.59 (0.10)0.30 (0.09)0.37 (0.08)60.71 (0.10)0.79 (0.13)0.95 (0.12)*
Nimesulide (1.0) 70.52 (0.07)0.26 (0.05)0.36 (0.04)60.70 (0.12)0.66 (0.13)0.87 (0.14)
Nimesulide (3.0) 70.55 (0.08)0.21 (0.04)0.38 (0.03)*60.61 (0.12)0.68 (0.16)0.97 (0.15)
Vehicle 90.70 (0.08)0.32 (0.04)0.29 (0.05)70.76 (0.11)0.92 (0.15)1.01 (0.15)
NS-398 (0.3) 90.64 (0.07)0.31 (0.04)0.32 (0.04)80.66 (0.10)0.80 (0.13)0.84 (0.12)
NS-398 (1.0) 80.64 (0.09)0.31 (0.07)0.47 (0.09)*80.69 (0.09)0.67 (0.11)0.95 (0.11)
NS-398 (3.0)100.76 (0.05)0.29 (0.05)0.47 (0.07)70.76 (0.09)0.78 (0.12)1.05 (0.13)
Vehicle 70.81 (0.12)0.36 (0.08)0.25 (0.05)70.70 (0.08)0.77 (0.12)0.78 (0.14)
Celecoxib (0.3) 70.81 (0.08)0.33 (0.06)0.27 (0.04)70.74 (0.10)0.74 (0.07)0.81 (0.08)
Celecoxib (1.0) 80.78 (0.08)0.34 (0.06)0.39 (0.06)70.67 (0.12)0.75 (0.16)1.00 (0.12)
Celecoxib (3.0) 70.77 (0.13)0.34 (0.05)0.48 (0.08)70.62 (0.07)0.65 (0.11)0.87 (0.11)

Higher doses of i.v. indomethacin, naproxen, paracetamol, flurbiprofen, nimesulide, NS 398 and celecoxib induced a significantly greater BVC than that at the end of the first hour of infusion with acetic acid (within groups), and by contrast with the decrease in vehicle-injected rats (Table 1). The increase in BVC was generally dose-dependent and, in some cases was >80% of that at the end of the first hour of acetic acid infusion. There was a significant difference from vehicle-injected rats (between groups), evaluated on ΔBVC recorded during the second hour of infusion with acetic acid, after all NSAIDs tested (Table 1). The ED50 calculated from the percentage change in injected groups corrected by the percentage change evaluated in the control group(s) is shown in Table 2. The peak MP was generally significantly decreased by the test compounds, but the decreases accounted for ≤ 30% of the decrease seen at the highest dose, and were generally not dose-dependent (data not shown).

Table 2.  Effect of the different compounds on cystometry in conscious rats 1 day after catheter implantation with bladders filled with dilute acetic acid, saline and in the isovolumic bladder contractions in anaesthetized rats
CompoundAcetic acid ED50, µg/kgSaline ED30, µg/kgVoidings Contractions DT10, µg/kg
  1. NC, not calculable: poor dose–response relationship; NT, not tested.

Aspirin27 681NC 2110
Indomethacin 1 487   841 590
Flurbiprofen    118   250  19
Naproxen   744   479 100
Meloxicam   594NC 208
Nimesulide   790   169 195
RofecoxibNTNT 135
NS 398 1 029 2 532  34
L 745337NTNT1131
Celecoxib 1 258 1 655  44
Paracetamol47 33837 0723296

Saline infusion for 1 h after stabilization of the urodynamic variables did not generally change BVC (Table 1), except in rats injected with flurbiprofen 0.3 mg/kg. Injections of vehicle i.v. did not significantly change BVC values during the second hour of bladder infusion with saline. All test compounds induced a dose-dependent increase in BVC, except for aspirin and meloxicam. Because the BVC was higher (than that at the end of acetic-acid infusion) before exposure to the compounds, the percentage increase after administration was generally ≤ 60%. Thus, the ED30 was calculated and shown in Table 2. Peak MP was again generally significantly but not dose-dependently decreased by administration of the test compounds (data not shown).

The BVC in rats at 5 days after catheterization were markedly higher than in rats 1 day after surgery. As shown in Table 3, the BVC was 1.57–1.79 mL, while in Table 1 it was 0.39–0.98 mL. By contrast there were no substantial differences in MP (data not shown).

Table 3.  Saline experiments 5 days after surgery; the effect of i.v. administration of naproxen and nimesulide on BVC in conscious rats. Data are the mean (sem) BVC in basal I and the mean recorded in the 1-h period after administration of vehicle or the test compounds
Treatments (doses in mg/kg)N ratsBasal IAfter
  • *

    P < 0.01 vs basal I.

Vehicle 91.78 (0.16)1.72 (0.12)
Naproxen (3) 91.57 (0.24)1.64 (0.24)
Vehicle121.79 (0.13)1.82 (0.12)
Nimesulide (3)121.65 (0.13)1.83 (0.12)*

Naproxen and nimesulide (both 3 mg/kg i.v.) were tested in rats 5 days after surgery. The increase in BVC induced by these compounds was less at 5 days than at 1 day after catheterization and not significantly different from the changes in vehicle-injected rats (Figs 1–3; Table 3). In Figs 2 and 3, the effect of naproxen and nimesulide, either in terms of percentage change or ΔBVC calculated from the basal value, was compared with their effect in rats after 1 day of catheterization and during infusion of the bladder either with saline or acetic acid. Figures 2 and 3 also show the percentage reduction in MP. The i.v. administration of 3 mg/kg of naproxen and nimesulide significantly decreased MP during saline infusion of the bladder in rats catheterized at 1 and 5 days, and in rats with bladders infused with acetic acid.

Figure 1.

Representative cystometrograms (and corresponding micturition volume traces) showing the effect of naproxen (3 mg/kg i.v.) in rats at 1 day (top cystometrogram) or 5 days (bottom cystometrogram) after surgery and with the bladder infused with saline. *injection of compound.

Figure 2.

Effect of increasing i.v. doses (0.3–1 and 3 mg/kg – dashed bars) of naproxen and vehicle (open bars) in the cystometry model in rats at 1 day after catheterization and with bladder infused with acetic acid (blue bars) or saline (black bars). Data obtained after naproxen (3 mg/kg i.v.) in rats 5 days after catheterization and with bladder infused with saline (red bars) are also reported. Data were expressed as percentage change (vs basal values) of BVC (top), as Δ mL of BVC (middle), and percentage change of MP (bottom). *P < 0.05 and **P < 0.01 vs basal values. Significance of Δ mL of BVC was vs vehicle-injected group.

Figure 3.

Effect of increasing i.v. doses (0.3–1 and 3 mg/kg – dashed bars) of nimesulide and vehicle (open bars) in the cystometry models used. Other details as in Fig. 2.

In anaesthetized rats with bladders infused with saline to give a constant volume, i.v. or i.c.v. injection of vehicle generally did not change the frequency of bladder contractions. Increasing doses of all the test compounds induced a dose-dependent disappearance of bladder contractions and a DT of 2–>35 min, depending on the compound administered (Table 4). The DT10 calculated from the DT data in Table 4 are given in Table 2. The amplitude of bladder contractions was not significantly affected by all the compounds tested. The i.c.v. administration of a very high dose of indomethacin and celecoxib (100 µg/rat) induced the disappearance of bladder contractions lasting only ≈ 6 min, similar or lower than that after i.v. administration of the corresponding dose.

Table 4.  The effect of the different compounds on isovolumic bladder contractions in anaesthetized rats
Compound, dose, µg/kg i.v.N ratsMean (sem) DT, min
  • *

    µg/rat i.c.v.

Aspirin
300 5 2.2 (1.2)
1000 5 5.0 (2.3)
3000 5 8.8 (1.0)
10000 519.9 (1.6)
30000 523.4 (4.9)
Indomethacin
300 6 5.1 (1.7)
1000 6 8.3 (3.3)
3000 835.0 (5.8)
100* 6 6.5 (2.4)
Flurbiprofen
10 6 6.1 (1.5)
30 6 11.4 (2.2)
100 625.0 (2.2)
300 536.1 (3.7)
Naproxen
100 6 8.6 (2.6 )
300 622.9 (4.9)
1000 629.9 (4.4)
Meloxicam
100 5 5.1 (1.6)
300 5 11.5 (4.0)
1000 522.5 (5.6)
3000 534.2 (13.0)
Nimesulide
100 5 5.6 (1.3)
300 6 11.4 (5.2)
1000 625.9 (7.7)
Rofecoxib
30 6 3.8 (1.0 )
100 6 11.9 (1.8)
300 516.4 (6.1)
NS-398
10 6 3.8 (0.6)
30 6 8.3 (2.9)
100 6 11.6 (1.9)
300 530.1 (7.4)
L 745337
100 4 3.0 (1.7)
300 4 5.1 (1.8)
1000 6 9.9 (2.4 )
Celecoxib
10 6 4.1 (0.7)
100 610.5 (2.0)
1000 628.4 (5.2)
100*10 6.2 (2.6)
Paracetamol
1000 4 3.7 (2.0)
3000 6 6.4 (2.3)
10000 619.4 (3.5)

The data summarized in Table 2 indicate that the NSAIDs tested are active in all cystometry models used. There was a close correlation (Fig. 4) between their potency at increasing BVC during cystometry in conscious rats 1 day after catheterization and with bladders infused with acetic acid vs bladders infused with saline (r = 0.890, P < 0.01). In addition, there was a significant correlation (r = 0.875, P < 0.01) between their potency at inhibiting the isovolumic bladder contractions in anaesthetized rats and increasing BVC during cystometry in conscious rats with the bladder infused with acetic acid (Fig. 4). There was a similar trend, although not significant (r = 0.595), when correlating the potency of the drugs in the isovolumic bladder contraction model with potency in the cystometry model with bladders infused with saline.

Figure 4.

Left: Correlation between the potency (log ED50) of some of the tested compounds on cystometry in conscious rats 1 day after catheterization and with bladder infused with acetic acid, and their potency (log ED30) on cystometry in conscious rats with the bladder infused with saline. Correlation coefficient = 0.89 (P < 0.01). Right: correlation between the potency (log DT10) of some of the tested compounds in the isovolumic bladder contractions model and their potency (log ED50) in increasing BVC in the cystometry in conscious rats 1 day after catheterization and with the bladder infused with acetic acid. In both figures the continuous line represents the regression line and the dotted line that of identity. Correlation coefficient = 0.875 (P < 0.01). Codes of the compounds: aspirin (A), indomethacin (B), flurbiprofen (C), naproxen (D), meloxicam (E), nimesulide (F), NS 398 (H), celecoxib (J), paracetamol (K).

In the present study we assessed the potency of compounds considered not selective for inhibiting the COX isozymes (aspirin, indomethacin, flurbiprofen, naproxen and paracetamol), and selective for COX-2 (meloxicam, nimesulide, rofecoxib, NS 398, L745337 and celecoxib). The potency of these compounds in our urodynamic models and their in vitro potency as inhibitors of COX isozymes were correlated to verify the relative importance of the different isozymes on the observed effects. Many in vitro assays are available to evaluate COX inhibitory property of compounds, and markedly variable results are produced by these systems [17]. To assess the correlation we therefore used the in vitro data reported by different authors [18–20], as shown in Table 5.

Table 5.  The potencies (IC50, nmol/L) of all compounds tested as inhibitors of prostanoid formation in different COX-1 and COX-2 assays
CompoundRiendeau et al. [19]*Vane et al. [18]Warner et al. [20]
COX-1COX-2COX-1COX-2COX-1COX-2
  • *

    Inhibition of PG E2 production in CHO cells stably transfected with human COX-1 and COX-2;

  • Inhibition of TxB2 production (COX-1) and PGE2 production (COX-2) in human whole blood;

  • Inhibition of TxB2 production in human whole blood (COX-1) and PGE2 in A549 (human airway epithelial) cells (COX-2).

  • ¶Inhibition of prostanoid production in bovine aortic endothelial cells (COX-1) and J774.2 macrophages (COX-2).

Aspirin    1 670278 0001 7007 500
Indomethacin    1826     28  1 680 13 130
Flurbiprofen     1.8 4   75 770
Naproxen    6226  9 30035 000
Meloxicam  1 810 6  4 800    4305 700 230
Nimesulide   780 9  9 200    52010 000 390
Rofecoxib    63 000 310
NS 398 1 900 6 16 800    1006 900 42
L 74533750 00060369 000  1 500>100 0001 300
Celecoxib   15 000     401 200 340
Paracetamol    >100 00064 000

The potency of the compounds in the isovolumic bladder voiding contractions correlated well with COX-2 inhibition, but not COX-1 (Fig. 5), regardless of the source of in vitro data used. Similarly, the potency in the cystometry model with bladder infused with acetic acid correlated better with the inhibitory potency on COX-2 than with that on COX-1 (Table 6).

Figure 5.

Correlation between the potency of some of test compounds as in vitro inhibitors of COX isozymes (pIC50; data from [18]) and their in vivo potency in the isovolumic bladder voiding contractions model (expressed as log DT10). Correlation coefficient for COX-2 = 0.882 (P < 0.01). Codes of the compounds: L 745337 (I); other codes as in Fig. 4.

Table 6.  Correlation between the potency (log DT10 or log ED50, µg/kg) of the different compounds tested in the isovolumic bladder voiding contractions in anaesthetized rats and in the cystometry model with the bladder infused with dilute acetic acid, with their potency (pIC50) as in vitro inhibitors of COX-1 and COX-2 isozymes. Data are the correlation coefficients (r) of the regression and (P) (no value = not significant)
ModelCOX-1COX-2COX isozymes data from:
Isovolumic bladder voiding contractions in anaesthetized rats0.5270.798 (<0.05)Riendeau et al. [19]
slope >10.882 (<0.01)Vane et al. [18]
slope >10.518Warner et al. [20]
Cystometry in conscious rats: bladder infused with acetic acid0.5520.646Riendeau et al. [19]
slope >10.876 (<0.05)Vane et al. [18]
0.1290.559Warner et al. [20]

DISCUSSION

Bladder afferents in the pelvic nerve, particularly in the rat, consist of tension receptors, volume receptors and capsaicin-sensitive C-fibres, including nociceptors [21]. PGs are involved in the physiological regulation of micturition during reflex activation of the urinary bladder. Endogenous PGs are produced locally after distension of the bladder wall and modulate the afferent branch of reflex micturition, lowering the threshold for eliciting voiding contractions [1,5–7] by acting through capsaicin-sensitive afferents [2]. It is likely that capsaicin-sensitive bladder nerves represent a target of PGs produced by urinary bladder, as: (i) nonselective COX inhibitors increase BVC and the pressure threshold for activating the micturition reflex [1,5,]; (ii) the impairment of these nerves, by pre-treating adults rats with high doses of capsaicin, produces urodynamic changes superimposable on those induced by COX inhibitors [2,5]; (iii) the effect of COX inhibitors is not apparent in animals pre-treated with capsaicin [2,5]. Thus, COX present in the urinary bladder may play a role in modulating bladder function.

The relative importance of the two COX isozymes in the modulating micturition reflex is poorly studied. There are two distinct COX isoforms, produced by different genes, and subject to differential regulation. COX-1, the constitutive isoform, is expressed in many tissue sites under basal conditions and has generally been assigned to various housekeeping functions, such as platelet aggregation, regulation of renal transport function and gastric mucosal cytoprotection. COX-2 is generally considered an inducible isoform; it has a limited pattern of basal expression but is rapidly induced by several stimuli, and is thought to play a role in inflammatory and proliferative responses [18]. However, COX-2 is constitutively expressed in neurones within the brain, and its expression is markedly and transiently up-regulated in neurones in response to excitatory stimuli [22,23], as well as in the bladder where it is stimulated during inflammation or bladder obstruction [12,13]. Furthermore, the distribution of COX-2 in the CNS and spinal cord, suggests that it is involved in processing and integrating visceral and special sensory input, and in elaborating the autonomic and behavioural responses [24,25].

Lecci et al. [14] evaluated the potency of a COX-2 selective (NS 398) and a nonselective (dexketoprofen) COX inhibitor in the induction of urodynamic changes in inflamed and normal bladders in rats. The effects of both NS 398 and dexketoprofen during cystometry were significantly greater in rats with bladder cystitis induced by lipopolysaccharide endotoxin or cyclophosphamide than in control rats. Similar results were obtained by other authors studying the urodynamic effects of indomethacin in controls or xylene-treated rats [26]. By contrast, NS 398 was much less potent than dexketoprofen in normal rats, and so it was suggested that the blockade of COX-1 is involved in urodynamic changes in control conditions, whereas COX-2 is mainly involved in inflamed bladder [14].

To verify this, we examined several nonselective COX inhibitors and compounds considered selective for COX-2 in different cystometry models. The results show that the potency of several nonselective or COX-2 selective compounds for inhibiting bladder overactivity induced by infusing the bladder with acetic acid, and their effect on BVC in rats with bladders infused with saline 1 day after catheterization, are substantially the same. These data are consistent, considering that tissue injury such as bladder catheterization during preparation for cystometry might increase COX expression and promote PG production in the bladder [27,28]. Cystometrographic recordings in conscious rats 1 day after catheterization and during saline infusion of the bladder confirmed that BVC was significantly less than at 5 days after catheterization, as previously reported [27]. The present data also indicate that the COX inhibitors naproxen and nimesulide were more active in enhancing the BVC of rats catheterized for 1 day, and less potent in rats 5 days after surgery, confirming the previously cited observations [14,26].

In the present experiments, the NSAIDs tested induced a similar reduction in MP either in irritated (by surgery or acetic acid infusion) and normal bladder (i.e. in rats at 5 days after surgery). It was shown that PGs contributed to the basal tone of the detrusor and modulated the activity of efferent bladder nerves. Prejunctional PG receptors on cholinergic nerve terminals of animals and man have been related to the potentiating effect of PGs on the electrically induced twitch response of isolated urinary bladder [3]. The similar effect on MP in normal and irritated bladder could be related to an effect at the level of efferent nerve discharge, independently of the degree of irritation.

In mice, several NSAIDs inhibit the production of PGE2 in a dose-related manner at doses that correlate well with both their potency to inhibit mouse brain COX and their anti-nociceptive potency in a mouse abdominal constriction test [29]. This central activity could be part of the inhibitory effect exerted by the tested compounds in the isovolumic bladder contraction model in anaesthetized rats. In these rats the bladder is catheterized via the external urethra with no surgery, and a marked irritation of the bladder may be excluded. However, when indomethacin and celecoxib were injected i.c.v. their effect was small, considering the very high dose administered. These results suggest that COX inhibitors mainly modulate afferent pathways in the bladder. The higher doses of the compounds needed to obtain a positive effect in conscious rats, compared with the anaesthetized ones, could be due to the need to inhibit also the peripheral effect due to bladder irritation and therefore a higher level of PGs.

Despite the marked differences in potency as inhibitors of the two COX isozymes reported for the tested anti-inflammatory compounds [18–20], there was generally a good correlation between the effect on the urodynamic models used and the COX-2 inhibiting potency, but not with COX-1 inhibition. Apart from the significant correlations, the results shown in Table 2 for the potency of, e.g. indomethacin and nimesulide in the urodynamic models, support the above assumption. The inhibitory activity (IC50) of these two compounds for COX-1 and COX-2 reported by Vane et al. [18] were 9.2 and 0.52 µmol/L for nimesulide and 0.028 and 1.68 µmol/L for indomethacin. Despite the marked difference of potency in inhibiting COX-1 (but similar potency for COX-2), these two compounds were similarly active in all the cystometry models.

The existence of a third COX isozyme, COX-3, was recently reported [30]. This enzyme is selectively inhibited by analgesic/antipyretic drugs such as paracetamol, and potently inhibited by some NSAIDs. As COX-3 is expressed at the site of inflammation and in the CNS, it was suggested that inhibiting this isozyme could represent a primary central mechanism by which these drugs decrease pain. However, the overall results obtained by studying the inhibitory potency of several anti-inflammatory compounds on all the COX isozymes indicate that COX-3 has activity that differs pharmacologically from COX-1 and COX-2, but is more similar to COX-1 [30], therefore excluding a relevant role of COX-3 in micturition reflex.

In conclusion, our results show that both nonselective and COX-2 selective inhibitors are more active in inhibiting the micturition reflex in rats with bladder overactivity due to bladder irritation than in normal rats. That the potency of the anti-inflammatory compounds tested in inhibiting bladder overactivity induced by chemical or surgical irritation, and that their activity in a cystometric model practically independent from bladder irritation (isovolumic bladder contractions in anaesthetized rats) is related to the potency as inhibitors of COX-2, suggests that the involvement of PGs in the micturition reflex in rats is mainly mediated via this isozyme.

In double-blind controlled studies in patients with detrusor instability, the nonselective COX inhibitors indomethacin and flurbiprofen had favourable effects, although they did not completely abolish detrusor overactivity [10]. Unfortunately, clinical studies showing evidence of therapeutic benefit of selective COX-2 inhibitors in patients with bladder disorders have not yet been published and therefore this represents the limit of the present findings.

CONFLICT OF INTEREST

None declared.

Ancillary