Effects of intravesical prostaglandin E₂ on bladder function are preserved in capsaicin-desensitized rats

Abstract

Prostaglandin E₂ (PGE₂) instilled into the bladder generates symptoms of urinary urgency in healthy women and reduces bladder capacity and urethral pressure in both humans and female rats. Systemic capsaicin desensitization, which causes degeneration of C-fibers, prevented PGE₂-mediated reductions in bladder capacity, suggesting that PGE₂ acts as an irritant (Maggi CA, Giuliani S, Conte B, Furio M, Santicioli P, Meli P, Gragnani L, Meli A. Eur J Pharmacol 145: 105–112, 1988). In the present study, we instilled PGE₂ in female rats after capsaicin desensitization but without the hypogastric nerve transection that was conducted in the Maggi et al. study. One week after capsaicin injection (125 mg/kg sc), rats underwent cystometric and urethral perfusion testing under urethane anesthesia with saline and 100 µM PGE₂. Similar to naïve rats, capsaicin-desensitized rats exhibited a reduction in bladder capacity from 1.23 ± 0.08 mL to 0.70 ± 0.10 mL (P = 0.002, n = 9), a reduction in urethral perfusion pressure from 19.3 ± 2.1 cmH₂O to 10.9 ± 1.2 cmH₂O (P = 0.004, n = 9), and a reduction in bladder compliance from 0.13 ± 0.020 mL/cmH₂O to 0.090 ± 0.014 mL/cmH₂O (P = 0.011, n = 9). Thus, changes in bladder function following the instillation of PGE₂ were not dependent on capsaicin-sensitive pathways. Further, these results suggest that urethral relaxation/weakness and/or increased detrusor pressure as a result of decreased compliance may contribute to urinary urgency and highlight potential targets for new therapies for overactive bladder.

INTRODUCTION

The critical symptom of overactive bladder is urinary urgency, defined as “a sudden and compelling desire to void that is difficult to defer” (1). Despite the prevalence of urinary urgency, the underlying pathophysiology remains poorly understood. A central hypothesis explaining urgency is the hypersensitization of bladder afferent pathways resulting in activation of normally silent C-fibers. Herein, we demonstrate that reductions in bladder capacity in female rats following intravesical instillation of prostaglandin E₂ (PGE₂) still occurred following degeneration of C-fiber afferents, implicating other mechanisms as contributing to urinary urgency.

Intravesical PGE₂ in healthy women causes a “strong desire to void” at reduced bladder volumes (2), and, in addition to reducing bladder capacity in female rats, intravesical PGE₂ increased external urethral sphincter (EUS) activity during bladder filling (3). We hypothesized that this increase in EUS activity might reflect striated muscle compensation for PGE₂-induced urethral smooth muscle relaxation, and intraurethral PGE₂ did indeed decrease urethral pressure, which occurs by elevation of cAMP levels in urethral smooth muscle (4). Additional testing suggested that the PGE₂-mediated reduction in bladder capacity was likely dependent on bladder filling, resulting in distension of the proximal urethra, which may contribute to urinary urgency (3). These observations suggest that one form of urinary urgency may be caused by increased physical distension of the proximal urethra, resulting from urethral relaxation or weakness and/or a decrease in bladder compliance (which was also observed following PGE₂ instillation).

However, previous work by Maggi and colleagues (5) suggested that PGE₂-mediated reductions in bladder capacity in rats were driven by capsaicin-sensitive bladder afferents. Following systemic capsaicin administration, which is known to impair C-fiber function (6), intravesical PGE₂ no longer reduced bladder capacity. These results led to the suggestion that intravesical PGE₂ is a bladder irritant, similar to the commonly used acetic acid model of overactive bladder (7). However, a bilateral sympathectomy (hypogastric nerve transection) was conducted before assessing the effects of PGE₂. Although 10 µM PGE₂ was sufficient to facilitate rhythmic bladder contractions when applied to the serosal surface of the bladder, when applied intravesically, 10 µM PGE₂ caused no change in bladder capacity. However, following hypogastric nerve transection, the rationale for which was not explained, intravesical administration of 10 µM PGE₂ did reduce bladder capacity. This reduction was subsequently inhibited following capsaicin desensitization. However, the impact of the nerve transection on the authors’ conclusions remains unclear.

We sought to determine whether systemic capsaicin desensitization would prevent PGE₂-mediated decreases in bladder capacity and urethral pressure in the absence of nerve transections. If systemic capsaicin desensitization prevented a reduction in bladder capacity, it would provide further support to the idea that PGE₂ irritates the bladder. If desensitization did not prevent decreases in bladder capacity and urethral pressure, it would further implicate urethral distention as contributing to urinary urgency.

METHODS

All animal care and experimental procedures were reviewed and approved by the Duke University Institutional Animal Care and Use Committee. Data come from 86 female Wistar rats, 44 from a previously published study (3) [median age and range: 18 (13–41) wk; median weight and range: 273 (232–360) g] and 42 additional animals [median age and range: 19 (16–47) wk; median weight and range: 279 (246–387) g]. Experiments from the previous study were conducted by C. L. Langdale (n = 18), J. A. Hokanson (n = 23), and D. Degoski (n = 3, under the observation of J. A. Hokanson). Experiments in this study were conducted by J. A. Hokanson (n = 22), J. Brooks (n = 16, under the observation of J. A. Hokanson), and C. L. Langdale (n = 4). Results were plotted as a function of investigator as well as by month, with no indication that these differences impacted the statements made in this manuscript. The only other factor that clearly changed between studies was a transition to using paddle electrodes for recording the electromyogram (EMG) of the urethral sphincter instead of percutaneous wires. When examining changes in EMG between the two cohorts, only the paddle electrode data were analyzed, as no wire electrodes were used in the capsaicin-desensitized cohort.

Administration of Capsaicin and Testing Timeline

Animals were administered 125 mg/kg capsaicin (Sigma-Aldrich) in three injections (25, 50, and 50 mg/kg sc, 12 h apart). Capsaicin was mixed with 10% v/v Tween (Sigma), 10% v/v ethanol (Sigma), and 80% v/v saline for a working concentration of 20 mg/mL 3–4 days before injection and stored at 20°C. The solution was vortexed just before each injection. For injections, animals were anesthetized with isoflurane or sevoflurane (1%–3%), injected, and then placed under anesthesia again (generally less than 1%) for 30 min while being closely monitored. After recovery from anesthesia, animals were returned to their cages. Terminal acute experiments were conducted 7–10 days after the initial capsaicin injection.

Eye-wipe testing using a capsaicin concentration of 100 µg/mL, diluted from the previously prepared stock, was conducted to verify that the animals were desensitized to capsaicin. Two to four drops of the capsaicin solution were applied to one eye. All animals had a negative test, responding with at most one wipe or swat. In the majority of cases, eye-wipe testing was conducted 3 days after the first injection. In four cases, eye-wipe testing was conducted 7–10 days after the first injection on the day of acute testing. This time window is well within the 9-mo time window (maximum duration tested) where rats that have been treated with 125 mg/kg of capsaicin have been shown to exhibit reduced eye wiping (8).

Surgical Preparation and Equipment Setup

Methods were similar to those used previously (3). Rats were anesthetized with urethane (1.2 g/kg sc, supplemented as necessary), and body temperature was monitored using an esophageal probe and kept at 36°C–38°C using a water blanket. Heart rate and arterial blood oxygen saturation levels were monitored using a pulse oximeter (Nonin Medical Inc., 2500 A VET). After exposing the bladder via an abdominal incision, a PE-90 catheter (427420, BD, Franklin Lakes, NJ) was inserted into the bladder and secured. For experiments involving only cystometric testing, the abdomen was sutured closed. For experiments with urethral perfusion, the abdomen was left open and covered with a piece of plastic wrap. The catheter was connected to a pressure transducer (ArgoTrans, Argon Medical Devices Inc., Plano, TX) and pump (Harvard Apparatus PHD 4400). Pressure signals were amplified and filtered (100 Hz low-pass, 13-6615-50, Gould Instruments, Valley View, OH) and sampled at 1 kHz using PowerLab (AD Instruments, Colorado Springs, CO). External urethral sphincter (EUS) EMG was measured using a bipolar paddle electrode placed underneath the pubic symphysis. Needle electrodes were used to record EUS EMG in a subset of the original experiments (3). EUS EMG leads were connected through a preamplifier (HIP5, Grass Products, Warwick, RI) to an amplifier (P511, Grass Products) with a subcutaneous needle as ground. Signals were filtered (3 Hz–3 kHz) and sampled at 20 kHz using PowerLab.

For experiments in which urethral perfusion testing was performed, a 4-0 or 5-0 silk suture was placed around the dorsal aspect of the urethra, and both ends were left to float without being tied (Fig. 1). After initial cystometric testing (saline only), the catheter was moved from the dome of the bladder to the urethra, and this suture was used to fix the catheter into the proximal urethra. The knot at the dome of the bladder was removed, so that any subsequent endogenous bladder filling (diuresis) emptied into the abdominal cavity, and plastic wrap was again placed over the open abdomen.

Figure 1.

Photos documenting the placement of suture to ligate the urethra and placement of the electrode to record EUS EMG. Photos are taken from different animals. A: zoomed-out view showing the exposed bladder and urethra. Skin covering the pubic bone is visible at the bottom of the image. B: close-up view of the pelvic nerve innervating the bladder and urethra. The suture is placed as close as possible to the urethra, between the large blood vessel and the ureter. The point of insertion is marked with an “x.” C: similar view as (B) in another animal. In this animal, there is more space for placement of the suture; however, the suture is placed as close as possible to the blood vessel. D: view with the bladder pushed caudally so that the ventral-most surface of the vagina is visible. In addition, a sheet of nervous tissue and blood vessels can be seen coming into the bladder (left side, below applicator). The suture is placed through this thin sheet and then around the dorsal aspect between the bladder/urethra and the vagina. E: photo showing a 5-45 forceps reaching through the thin sheet of tissue to grab the suture. In this experiment, the EMG electrode (wires at the bottom of the image) were placed prior to suture placement. F: in this photo, the suture is in place and ready to ligate the urethra after the catheter is inserted into the proximal urethra. The EMG electrode being positioned underneath the pubic bone. After placement, the EMG wires were bent toward the tail and sutured to the skin above the pubic bone. b, bladder; EMG, electromyogram; EUS, external urethral sphincter; ua, urethra; ur, ureter; us, uterus; v, vagina.

Testing

The bladder was filled continuously with an open urethra for at least 45 min with room temperature saline (2–8 mL/h) to allow postsurgical recovery. The bladder was subsequently emptied, and three to six single-fill cystometric trials (CMGs) were recorded. For each CMG, the bladder was filled until a micturition event or leaking (e.g., overflow incontinence) was observed, at which time the infusion pump was turned off. After the bladder pressure returned to a steady-state baseline or 1 min had passed, the bladder was emptied via the catheter. Voided and residual volumes were recorded and used to calculate bladder capacity and voiding efficiency. The bladder remained empty for 2–3 min before the next trial. For experiments without urethral infusion, PGE₂ was subsequently instilled into the bladder. Similar to saline, PGE₂ was infused continuously for at least 45 min, followed by three to six single-fill CMGs.

Urethral perfusion trials consisted of perfusing the urethra at 1 mL/min for 2 min, followed by 6 min at 0.005 mL/min. PGE₂ trials immediately followed the saline trials. Typically, five to six trials were conducted with each perfusion medium. After all the trials were collected, the animal was euthanized with Euthasol, and an additional trial was collected.

In nine capsaicin-naïve animals, saline cystometry was followed with PGE₂ cystometry at increasing concentrations of PGE₂ (30, 60, and 100 µM). Similar to other cystometric testing, 45 min of infusion at each concentration was followed by two or more single-fill CMGs.

In six capsaicin-naïve animals, saline or PGE₂ was applied to the external bladder (serosal) surface near the bladder dome. Five drops of saline or PGE₂ at various concentrations (5–100 µM), warmed to 37°C, were applied to the bladder at either 50% of bladder capacity (five experiments) or one-third and two-third of bladder capacity (1 experiment). Bladder pressure was monitored for 5 min (5 experiments) or 10 min (1 experiment) after application or until a reflexive bladder contraction occurred. In the last four experiments, bladder filling began after this waiting period until voiding to evaluate changes in bladder capacity. In the first two experiments, bladder capacity was computed for trials in which application of PGE₂ led to a bladder contraction—saline never led to a bladder contraction. In the first two experiments, two PFA-coated platinum-iridium wires (0.0055-in. diameter, A-M Systems) were inserted percutaneously (via needle) on each side of the urinary meatus and lodged close to the urethral sphincter to record EUS EMG. In the last four experiments, a paddle electrode was used to record EUS EMG.

PGE₂ (Sigma-Aldrich and Tocris) was dissolved in ethanol at 10 mM concentration and stored in a −20°C freezer. On experiment days, a portion of this stock solution was diluted in saline to 100 µM.

Data Analysis

Bladder capacity was calculated as the sum of the voided and residual volumes. Voiding efficiency was calculated as the voided volume divided by the bladder capacity. Bladder compliance was calculated by dividing the fill rate by the slope of bladder pressure versus time. This slope was estimated using Thiel–Sen regression, which is less sensitive to outliers (such as from nonvoiding contractions), to fit a line to the bladder pressure trace during filling between 20% and 80% of the total filling time (3). Average rectified EMG activity was calculated over the period between 75% and 95% of the total filling time and normalized to the average rectified value during the first 10% of the total filling time in the same trial. The static urethral perfusion pressure was calculated as the average of the middle 60% of the period of low-rate urethral perfusion (between 1.2 and 4.8 min) captured after the high-filling rate perfusion. For each condition, the values measured during the last three trials, after the system had stabilized, were averaged together and used for statistical testing.

For experiments where PGE₂ was applied to the serosal surface of the bladder, a ratio of bladder pressures was calculated to evaluate the increase in bladder pressure due to application of the PGE₂ (or saline). This ratio was calculated as the pressure at the last 50% of the waiting period after serosal application—either until the pump started again, the start of a bladder contraction, or trial end—relative to the pressure at the time of serosal application (30 s centered on the application event). In the one experiment in which PGE₂ or saline was applied twice during a fill, the latter application was analyzed. The resulting pressure ratios were averaged for each experimental condition. Statistical comparisons between means were conducted using either paired t tests for data before and after PGE₂ administration or unpaired t tests for data from capsaicin-naïve versus capsaicin-desensitized animals. P values <0.05 were considered statistically significant. All computations were performed in MATLAB (MathWorks). The analyses between capsaicin-desensitized animals and capsaicin-naïve animals assume that the age or weight of the animals did not impact the outcome measures, which on visual inspection appeared to be the case. In addition, multivariate linear regression models were created with age, weight, and capsaicin status as factors in GraphPad (8.4.3) to confirm that the observed differences were not simply a function of age or weight differences.

Analyses of changes in bladder capacity from dose-response or serosal experiments were analyzed using mixed-effects analysis in GraphPad (8.4.3). Prior to analysis, bladder pressure data were normalized and then log transformed to reduce variance and skew. Post hoc comparisons with saline controls were made using the Dunnett’s multiple-comparisons correction method. P values <0.05 were considered statistically significant. Four animals died of apparent respiratory failure (shallow and infrequent breathing) 5–10 min after the second capsaicin injection (first 50 mg/kg dose) (8). A fifth animal showed strong respiratory distress during the second injection and survived, but that animal died after the initial dose of urethane. In two animals in which dose-response data were evaluated, control bladder capacities were abnormally low (0.17 and 0.24 mL). In these experiments, no change in bladder capacity was noted after intravesical PGE₂ administration, even at 100 µM, and these two experiments were excluded from analysis. All other animals completed planned testing, and their data were included in the analysis.

RESULTS

Example CMGs from capsaicin-desensitized rats are shown in Fig. 2A. In three of the 18 capsaicin-desensitized animals that underwent cystometry, no reflex contraction was observed, and subsequently overflow incontinence occurred. Capsaicin-desensitized rats had larger bladder capacities (1.23 ± 0.08 mL, means ± SE, n = 18) than capsaicin-naïve rats (0.67 ± 0.05 mL, n = 49, P < 0.0001, Fig. 2B). Voiding efficiency was also decreased in capsaicin-desensitized animals compared with capsaicin-naïve controls, from 36.9 ± 2.2% to 24.8 ± 2.7% (P = 0.0052, Fig. 2C, n = 49 naïve, n = 15 capsaicin-desensitized, animals with overflow incontinence excluded). We did not observe differences in EUS EMG activity during filling (P = 0.93, n = 26 naïve, animals with wire EMG electrodes excluded, n = 18 capsaicin-desensitized, Fig. 2D), bladder compliance (P = 0.43, n = 49 naïve, n = 18 capsaicin-desensitized, Fig. 2E), or urethral perfusion pressure (P = 0.53, n = 7 naïve, n = 9 capsaicin-desensitized, Fig. 2F) between capsaicin-naïve and capsaicin-desensitized animals. The same results held when examining the outcome measures as a function of capsaicin status, age, and weight using multivariable regression. Capsaicin status impacted bladder capacity and voiding efficiency, but age and weight were never significant factors.

Figure 2.

Effects of capsaicin desensitization on cystometric parameters and urethral perfusion pressure. A: example cystometrograms from two animals after capsaicin desensitization. The top traces are bladder pressure, and the bottom traces are urethral sphincter EMG. In three of the 18 experiments, overflow incontinence occurred (right). Capsaicin desensitization increased bladder capacity (B; BC; P < 0.0001, n = 49 naïve, n = 18 capsaicin-desensitized) and decreased voiding efficiency (C; VE; P = 0.0052, n = 49 naïve, n = 15 capsaicin-desensitized, three omitted due to overflow incontinence). There was no change in urethral sphincter EMG (D; P = 0.93, n = 26 naïve, n = 18 capsaicin-desensitized, naïve animals with wire electrodes were excluded), bladder compliance (E; P = 0.43, n = 49 naïve, n = 18 capsaicin-desensitized), or urethral perfusion pressure (F; P = 0.53, n = 7 naïve, n = 9 capsaicin-desensitized) between the two groups. Data in the naïve group (B–F) are from a previous publication (24) with additional experiments (n = 3) for the urethral perfusion group (cystometry data were also collected) and additional PGE₂ dose-response and serosal experiments (n = 13). The additional urethral perfusion data were collected to increase power and to ensure that the techniques and readouts were comparable with previously collected data. Animals for (C–E) are the same as (B). Box plots were created using the boxplot() command in MATLAB (2019 b). Medians are indicated with a line and boxes span from the 25th to 75th percentiles. Whiskers extend to the most extreme data points not considered outliers. Outliers are indicated with “+” symbols (for visualization only, all data were included in the analyses). Two-tailed unpaired t tests were used to compare means between the two groups. EMG, electromyogram.

Previous results indicate that PGE₂ decreases bladder capacity and urethral perfusion pressure in capsaicin-naïve animals (3). Of primary interest in the present study was whether PGE₂ also reduced bladder capacity and urethral perfusion pressure after capsaicin desensitization. Figure 3, A and B, shows example CMGs from saline and PGE₂ trials in two different experiments. In both examples, bladder capacity decreased, and in the second example, leaking occurred. PGE₂-induced leaking occurred at low pressures (8–15 cmH₂O, mean: 12 cmH₂O, n = 5) compared with pressures in the three previously mentioned animals that demonstrated overflow incontinence before PGE₂ (31, 33, and 39 cmH₂O, experiment averages). Four animals were from a cohort that underwent monitoring in metabolic cages before undergoing acute testing at 43 wk of age. All four of these animals leaked with PGE₂, one of which leaked before PGE₂ administration. In a separate group of five animals that were 18 wk of age, only one leaked with PGE₂, and none leaked before PGE₂ administration.

Figure 3.

Data from capsaicin-desensitized animals during saline cystometry or filling with 100 µM PGE₂. A: example bladder pressure (top) and EUS EMG (bottom) traces during saline (left) and PGE₂ filling (right) in one animal. PGE₂ decreased bladder capacity without a meaningful change in voiding efficiency. B: example traces from another animal. In addition to decreasing bladder capacity, PGE₂ led to leaking at low pressure (experiment average: 15 cmH₂O). C: filling with PGE₂ decreased bladder capacity (P = 0.002, n = 9). D: voiding efficiency did not change after PGE₂ (P = 0.57, n = 4, five animals removed due to leakage). E: bladder compliance decreased after PGE₂ (P = 0.011, n = 9). F: EMG activity did not change after PGE₂ (P = 0.24). G: example traces of urethral perfusion pressure (top) and EUS EMG activity (bottom) for five trials. Perfusion of PGE₂ through the urethra led to a reduction in urethral pressure. Measurement of the pressure after the animal was euthanized allowed for the assessment of passive pressure. H: urethral perfusion pressure decreased following PGE₂ perfusion (P = 0.004, n = 9). Percent changes for (H) were scaled such that −100% corresponds to a decrease to the pressure measured after euthanasia. Statistical comparisons were conducted using paired t tests. Square data points (n = 9) in C–F indicate data from capsaicin-desensitized animals that underwent saline cystometry followed by urethral perfusion testing (H). EMG, electromyogram; EUS, external urethral sphincter; PGE₂, prostaglandin E₂.

PGE₂ reduced bladder capacity in capsaicin-desensitized rats from 1.23 ± 0.08 mL to 0.70 ± 0.10 mL (P = 0.002, n = 9 for paired t test, n = 18 for means ± SE in saline group, Fig. 3C). PGE₂ also reduced urethral perfusion pressure from 19.3 ± 2.1 cmH₂O to 10.9 ± 1.2 cmH₂O (P = 0.004, n = 9, Fig. 3H) and reduced bladder compliance from 0.13 ± 0.020 mL/cmH₂O to 0.090 ± 0.014 mL/cmH₂O (P = 0.011, n = 9 paired t test, n = 18 for saline means ± SE, Fig. 3E). No change in EUS EMG activity (P = 0.24, n = 9 paired t test, n = 18 for saline means ± SE, Fig. 3F) or voiding efficiency (P = 0.57, n = 4 paired t test, 5 omitted due to leakage, n = 15 for saline means ± SE, 3 omitted due to leakage, Fig. 3D) was observed. In addition to the experiments with PGE₂ cystometry (n = 9), data from nine additional capsaicin-desensitized animals that underwent saline cystometry followed by urethral perfusion testing were plotted in Fig. 3, C–F, for the saline condition (marked as squares; circles indicate animals that have both saline and PGE₂ data).

It was reported previously (5) that 10 µM PGE₂ was insufficient to influence bladder capacity in a neurally intact animal. Prior to our previous study (3), we tested different concentrations of PGE₂ in four animals, starting at 30 µM (n = 3) or 60 µM (n = 1). One animal appeared to respond to PGE₂ at 60 µM, another at 100 µM, and two failed to respond at any dose, including 100 µM. On closer examination, these two experiments had abnormally low average bladder capacities (0.17 and 0.24 mL) during saline cystometry (the next lowest value was 0.37 mL). Data from five additional animals were collected to evaluate changes in bladder capacity as a function of PGE₂ dose (Fig. 4A; P = 0.0016, mixed-effects model, n = 7). Only 100 µM PGE₂ generated a decrease in bladder capacity to a mean of 81% [(95% confidence interval: 70%–94%), P = 0.0162, n = 6] of control values. The study conducted in healthy women used 85 µM PGE₂ intravesically (2). The lack of effects at 30 µM and 60 µM highlights that 10 µM PGE₂ intravesically in an anesthetized rat is unlikely to generate a meaningful impact on bladder capacity without additional manipulations.

Figure 4.

Effects of dose and route of application of PGE₂. A: normalized bladder capacities from intravesical PGE₂ dose-response testing. Only 100 µM generated a decrease in bladder capacity (P = 0.0162, n = 6; 30 µM, P = 0.29, n = 7; 60 µM, P = 0.81, n = 7). Results were excluded from two experiments with abnormally low control bladder capacities (0.24 and 0.17 mL). In those experiments PGE₂ did not decrease bladder capacity at any dose. B: experimental setup for testing PGE₂ or saline on the outside surface of the bladder. C: example of cystometric trials where at 50% of expected bladder capacity (based on previous trials), the pump was stopped and either saline or PGE₂ was applied to the serosal surface of the bladder (five drops). If fluid application did not evoke bladder emptying after 5 min, the pump was started again to establish bladder capacity. At all tested PGE₂ doses, serosal application led to decreases in bladder capacity (D) [PGE₂ (5/10 µM), P = 0.029, n = 5; PGE₂ (30 µM), P = 0.0006, n = 4; and PGE₂ (100 µM), P = 0.0020, n = 4] and increases in pressure (E) [PGE₂ (5/10 µM), P = 0.0053, n = 6; PGE₂ (30 µM), P = 0.035, n = 5; and PGE₂ (100 µM), P = 0.0077, n = 4]. Statistical comparisons were conducted using mixed-effects models with post hoc testing compared with the saline control condition using Dunnett’s multiple-comparisons method. PGE₂, prostaglandin E₂.

In six animals, PGE₂ was applied to the serosal surface of the bladder at various levels of bladder fullness (Fig. 4B) to evaluate changes in bladder pressure and bladder capacity as a function of PGE₂ dose. Figure 4C shows an example experiment in which application of PGE₂ led to increases in bladder pressure, eventually leading to bladder contractions, and a subsequent reduction in bladder capacity. All tested doses of PGE₂ applied to the serosal surface of the bladder resulted in a decrease in bladder capacity (P = 0.0009, mixed-effects model, n = 5). Average values were 0.82 ± 0.13 mL (saline), 0.58 ± 0.05 mL [PGE₂ (5/10 µM), P = 0.029, n = 5], 0.34 ± 0.08 mL [PGE₂ (30 µM), P = 0.0006, n = 4], and 0.42 ± 0.07 mL [PGE₂ (100 µM), P = 0.0020, n = 4]. PGE₂ increased bladder pressure to mean values of 157% [95% confidence interval: 128%–191%, PGE₂ (5/10 µM), P = 0.0053, n = 6], 197% [95% confidence interval: 124%–313%, PGE₂ (30 µM), P = 0.035, n = 5], and 194% [95% confidence interval: 151%–250%, PGE₂ (100 µM), P = 0.0077, n = 4] of control values (P = 0.0173, mixed-effects model, n = 6). These results demonstrate the importance of the method of PGE₂ administration and highlight that administration of PGE₂ to the serosal surface of the bladder (or to bladder strips, as is common) is not equivalent to intravesical administration of PGE₂.

DISCUSSION

PGE₂ reduced bladder capacity in capsaicin-desensitized animals. We used a higher concentration of PGE₂ than Maggi et al. (5) (100 µM versus 10 µM), but we did not transect the hypogastric nerves. Similarly, Patra and Thorneloe (9) observed that intravesical 120 µM PGE₂ reduced the micturition interval in capsaicin-desensitized spontaneously hypertensive rats (SHR) using awake restrained cystometry. The reduction in micturition interval in capsaicin-desensitized animals (∼25%) was not as large as that observed in nondesensitized animals (∼50%), leading them to suggest that only a portion of the PGE₂ effect is via capsaicin-sensitive afferents. We saw the opposite: a larger decrease in capacity in capsaicin-desensitized animals (40.6 ± 9.0% mean decrease) as compared with nondesensitized animals from our previous study (16.7 ± 3.9% mean decrease, p = 0.010, unpaired t test) (3). Direct comparison between our study and theirs is difficult because of the use of different rat strains (SHR vs. Wistar) and comparing the micturition interval versus bladder capacity, but the results from both studies suggest that PGE₂ does not modulate the micturition reflex exclusively via capsaicin-sensitive afferents.

Overflow incontinence was noted in three of 18 capsaicin-desensitized animals undergoing saline cystometry. We have not previously observed overflow incontinence during standard saline cystometry in capsaicin-naïve animals, and this suggests that it is very likely that overflow incontinence is a result of the capsaicin desensitization, which itself could be characterized by a less active detrusor [more volume needed to initiate a contraction, i.e., increased bladder capacity (Fig. 2B), and a less efficient contraction, i.e., decreased voiding efficiency (Fig. 2C)]. Overflow incontinence was also observed by other investigators following 50 mg/kg capsaicin administration (instead of the 125 mg/kg used here) (10, 11).

Urethral pressure and bladder compliance were reduced in capsaicin-desensitized animals after PGE₂ administration, as well. We are not aware of other studies quantifying these PGE₂-mediated effects in capsaicin-desensitized animals. As described in our previous study (3) and below in Physical Model of Urinary Urgency. We expect that both of these changes would contribute to a reduction in bladder capacity.

PGE₂ Mode of Administration

Much of the initial work with PGE₂ was done on bladder strips, and it was found to increase bladder tone (12, 13). However, despite initial promising results (14), PGE₂ was not consistently effective at treating urinary retention (15, 16). Although it is possible that better patient selection would identify a subset responsive to PGE₂ treatment, it is likely that one aspect of failure in translating this therapy from animal testing to humans came from the differences in mode of administration (bladder strips vs. intravesical administration). We found a strong effect of topical application of PGE₂ to the serosal surface of the bladder (stronger contractions) at much lower concentrations (5 and 10 µM) compared with concentrations required for an effect when instilled intravesically (Fig. 4). It remains to be tested whether PGE₂ administered to the serosal surface of the urethra also shows differential responsiveness to dosing like those observed for the bladder.

The difference between intravesical and serosal PGE₂ suggests that PGE₂ diffusion across the urothelium is likely limited. Interestingly, only a small subset of patients given PGE₂ intravesically for retention complained of side effects, notably bladder discomfort (14–23). As PGE₂ has been shown to activate C-fibers after bladder urothelium disruption (via administration of protamine sulfate) (24), it is possible that the subset of patients in which PGE₂ produced discomfort had urothelial dysfunction. This difference in response between intact and disrupted urothelium would be similar to what has been observed with potassium, where potassium is an irritant only when urothelial integrity has been disrupted (25, 26). PGE₂ also activates some C-fibers and Aδ-fibers in a rat model with L6 and S1 dorsal root transections (27). Whether these transections impacted the urothelium is unclear. If these transections did not impact the urothelium, these results may highlight a component of the PGE₂ pathway that truly is capsaicin sensitive (something that cannot be excluded in this study). Alternatively, these changes may be secondary to changes in smooth muscle function (as discussed below in Model of PGE₂ During Bladder Filling).

Low Concentration PGE₂ and Hypogastric Nerve Transection

Given the discrepancy between our results and those of Maggi, it would be helpful to confirm that hypogastric nerve transection does indeed facilitate low-dose PGE₂, reducing bladder capacity but only in non–capsaicin-desensitized animals. However, it is unclear why hypogastric nerve transection in conjunction with low-dose PGE₂ only reduces bladder capacity in capsaicin-naïve animals.

Three potential explanations of why capsaicin prevented a PGE₂-mediated decrease in bladder capacity following hypogastric nerve transection are illustrated in Fig. 5. In the first explanation (Fig. 5A), the decrease in bladder capacity was due to the hypogastric nerve transection, not PGE₂, and capsaicin desensitization delayed or prevented this transection-mediated decrease. Effects of hypogastric nerve transection on bladder capacity are variable when evaluated shortly (∼15 min) after transection (28–30); however, reductions in bladder capacity are observed 3 days after transection (29). A more detailed time course of the decrease in bladder capacity is not known in rats, and it is possible that sequential testing led to a reduction in bladder capacity on the second test, independent of PGE₂ instillation.

Figure 5.

Possible explanations for why capsaicin impaired the effects of 10 µM PGE₂ only after hypogastric nerve (HGN) transection (Tx). A: capsaicin slows or prevents HGN Tx-mediated decrease in bladder capacity. B: HGN Tx increased potency of PGE₂, and capsaicin desensitization decreased potency. C: HGN Tx disrupted the urothelial barrier function giving PGE₂ an alternative pathway to reduce bladder capacity. PGE₂, prostaglandin E₂.

In the second explanation (Fig. 5B), hypogastric nerve transection amplified the physiological effects of PGE₂. Although the rationale for hypogastric nerve transection was not provided, it seems likely that they chose to transect the hypogastric nerves, as these nerves are viewed as promoting bladder storage (31, 32). However, it is not clear why PGE₂ would be more potent in a disinhibited system, or how capsaicin desensitization would decrease this potency.

Finally, in the third explanation (Fig. 5C), hypogastric nerve transection impairs urothelial function, such that PGE₂ more readily crosses the urothelial barrier to activate capsaicin-sensitive sensory neurons. Spinal cord injury impairs urothelial function within 2 h of injury (earliest tested time point, 33). To our knowledge, the impact of hypogastric nerve transection on urothelial dysfunction has not been examined. It is possible that hypogastric nerve transection similarly impairs the urothelial barrier function, allowing low concentrations of PGE₂ to activate bladder afferents. In this scenario, PGE₂-mediated decreases in bladder capacity would indeed occur via capsaicin-sensitive afferents that are subsequently inhibited by capsaicin desensitization. The mechanism by which spinal cord injury impacts urothelial function, as well as whether or not the hypogastric or pelvic nerves are involved (or if it is hormonal), warrants further investigation. The signaling properties of the urothelium are widely acknowledged; urothelium function is an important contributor to urinary tract function (34), and clarification of how the urothelium is regulated remains an important question.

Model of PGE₂ During Bladder Filling

Papers by Maggi and colleagues (5, 35) include a model to show factors, such as PGE₂, that regulate bladder capacity. In this model, bladder distension from filling directly triggers increased detrusor muscle activity, as well as the release of PGE₂, which also increases detrusor muscle activity. Bladder distension increases PGE₂ production (36–38), and one possible role for distension-evoked PGE₂ release is as a safety mechanism to prevent bladder overdistension (e.g., due to outlet obstruction) by increasing bladder tone and relaxing the urethra as the bladder fills.

We revised this model with two critical updates (Fig. 6A). First, we added the urethral muscle to the model. PGE₂ generates relaxation of the urethra in rats (3, 4) and humans (2, 17), and our current results (Fig. 3) suggest that this relaxation is not dependent on a capsaicin-sensitive pathway. Second, we expect that changes to the smooth muscle of the lower urinary tract will result in afferent activation, including capsaicin-insensitive afferents. For example, as the urethra relaxes, we expect increased activation of urethral afferents from fluid distension (39). In addition, the reduction in compliance, presumably due to bladder smooth muscle activation, would also be expected to result in increased capsaicin-insensitive afferent activation.

Figure 6.

Models of PGE₂-mediated activation of afferents and model of urinary urgency. A: modification of schematic from (5) to include the urethra, as well as afferent activity from capsaicin-insensitive sensory fibers. The original model, with some slight modifications, is highlighted in gray. PGE₂ facilitates bladder emptying by increasing detrusor smooth muscle activity and relaxing the urethra. In the previous model, capsaicin desensitization would prevent PGE₂-mediated afferent discharge to the CNS. Whereas in the modified model, capsaicin-insensitive afferents (including some but not all Aδ-fibers, as some Aδ-fibers are capsaicin insensitive) (41) respond to PGE₂-mediated changes in smooth muscle tone. We are not aware of studies that examined whether PGE₂ directly modulates capsaicin-insensitive neurons (i.e., via receptors on the neuron), as opposed to these neurons changing activation via other PGE₂-sensitive pathways (e.g., smooth muscle). B: physical model of urinary urgency. The healthy bladder is highly compliant and relaxed during bladder filling; the urethra is strong. Intravesical PGE₂ reduces compliance and causes urethral relaxation. In this model, urgency results from an impending sense of urine loss due to muscular changes. CNS, central nervous system; PGE₂, prostaglandin E₂.

Physical Model of Urinary Urgency

Intravesical PGE₂ causes urinary urgency when administered to women with urinary retention (14, 17) and healthy women (2). The work by Maggi et al. suggested this urgency was due to capsaicin-sensitive afferents, supporting the hypothesis that urinary urgency is the result of hypersensitization of bladder afferent pathways, including increased C-fiber activation. However, prostaglandins are known to be potent modulators of smooth muscle function (40), including urethral smooth muscle function (4). Our findings that a reduction in bladder capacity and urethral pressure still occurred in capsaicin-desensitized animals suggest that an alternative hypothesis or additional component is needed to explain urinary urgency.

Figure 6B demonstrates a physical model of urinary urgency, where increased mechanical pressure on the proximal urethra (from urethral relaxation, bladder activation, or both) leads to an impending sense of urine loss at lower bladder volumes (41). This physical model of urinary urgency suggests that strengthening or increasing activation of the urethra and/or increasing bladder compliance or reducing the basal tone of detrusor smooth muscle should improve urinary urgency symptoms. Other models of urinary urgency include the myogenic theory, the neurogenic theory, and the autonomous bladder theory (42). One aspect of the neurogenic theory is that an increase in C-fiber activity promotes urinary urgency. It is important to note that this may still be a component of the urinary urgency resulting from PGE₂ administration. However, hypogastric nerve transection prevented decreases in bladder capacity from intravesical acetic acid (43). In the Maggi et al.’s study, it was only after hypogastric nerve transection that 10 µM PGE₂ reduced bladder capacity. If PGE₂ was acting as an irritant like acetic acid, one might expect that hypogastric nerve transection would not have been sufficient for 10 µM PGE₂ to reduce bladder capacity and, in fact, would have prevented higher doses from having an effect on bladder capacity.

Although the myogenic theory of urinary urgency (44) and our physical model of urgency may overlap, there are important differences. The myogenic theory has focused on denervation of detrusor smooth muscle leading to increased excitability and coupling between cells and an increase in detrusor overactivity. Although denervation and detrusor overactivity may occur in the physical model, they are not critical components. When PGE₂, or the analog, sulprostone, was given to six healthy women, these drugs caused detrusor overactivity in only seven of the 12 fills (six women, two drugs). In contrast, urinary urgency was always associated with drug administration. The only changes with both drugs were reductions in bladder capacity and maximum urethral closure pressure. Further, the myogenic theory also omits reference to the urethra, whereas the urethra is a critical component of the physical theory.

Limitations and Possible Future Studies

The eye-wipe test is commonly used to assess for capsaicin desensitization, but it does not prove that capsaicin-sensitive neurons are nonfunctional throughout the body. Histology would be beneficial in demonstrating how capsaicin-sensitive neurons recover over time following systemic administration of capsaicin. In addition, transient receptor potential V1 (TRPV1) knockout mice (45) may be useful for further demonstrating the effects of PGE₂ (or possible lack thereof) without the receptors (TRPV1) that capsaicin is known to activate.

In our experiments, estrous cycle status was not tested. Estrous status is known to influence lower urinary tract function (46–48). Although the magnitude of the effects of PGE₂ on bladder capacity, urethral pressure, etc., may vary during different phases of the estrous cycle, we do not expect that incorporating estrous cycle status into our analysis would change our conclusions, as our effects were quite consistent; eight of nine capsaicin-desensitized rats exhibited a decrease in bladder capacity of 23% or more following PGE₂ and another eight of nine (different animals) demonstrated a decrease in urethral pressure of 34% or more. Nevertheless, estrous cycle status may modify the degree to which PGE₂ impacts lower urinary tract function.

Conclusions

PGE₂ instillation produced effects in capsaicin-desensitized animals that were similar to those observed in naïve animals, including reductions in bladder capacity, urethral pressure, and bladder compliance. Although effects of PGE₂ on capsaicin-sensitive afferents cannot be excluded, our results highlight that capsaicin-sensitive pathways are not required for PGE₂-mediated actions on bladder function. As PGE₂ causes a strong sense of urinary urgency in women, these results suggest a potential contribution of urethral dysfunction in urinary urgency (physical theory of urinary urgency). This may include urethral smooth muscle dysfunction, which is noticeably absent from current theories of urinary urgency. Targeting smooth muscle dysfunction, particularly urethral smooth muscle dysfunction and even urethral dysfunction more generally, may be a promising approach to treating urinary urgency and other overactive bladder symptoms.

GRANTS

This work was supported by the GlaxoSmithKline/Galvani Bioelectronics Research Program and in part by National Institute of Diabetes and Digestive and Kidney Diseases Grant K12DK100024.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.A.H., C.L.L., P.H.M., A.S., and W.M.G. conceived and designed research; J.A.H. and C.L.L. performed experiments; J.A.H. and C.L.L. analyzed data; J.A.H. and C.L.L. interpreted results of experiments; J.A.H. prepared figures; J.A.H. drafted manuscript; J.A.H., C.L.L., and W.M.G. edited and revised manuscript; J.A.H., C.L.L., P.H.M., A.S., and W.M.G. approved final version of manuscript.