Data were collected from adult female pigs (Sus scrofa domestica, 6 months) and rats (Wistar 200–240 g). Pig bladders were obtained from the local abattoir, immediately after slaughter (Langford, Bristol, UK). Rats were sourced from Charles River UK Ltd. and housed at the University of Bristol Animal Services Unit, where they were maintained at 22°C in plastic cages with bedding, under a 12/12-hour light-dark environment and with free access to food and water. Animal care and experimental procedures with rats were in compliance with the University of Bristol Ethics Committee and UK legislation under the Animals (Scientific Procedures) Act 1986 Amendment Regulations (SI 2012/3039) and adhered to ARRIVE guidelines [9].

The bladder, ureters, urethra, and distal abdominal aorta with the vesical branches were removed and immediately transported to the laboratory in ice-cold, gassed Tyrode’s solution. The aorta was cannulated with a 14-G needle, the redundant branches were ligated, and the bladder was perfused (10mL/min) with Tyrode’s solution by a peristaltic pump: drug interventions were introduced into the arterial perfusate. The preparation was put into a 20-L closed water-jacketed (36°C) chamber, into which venous outflow drained through a siphon, to maintain a humidified environment. A 6-channel double-lumen catheter was inserted into the bladder via the left ureter, secured with a Vicryl suture, and connected to a pressure transducer; the right ureter was ligated. The bladder was filled with Tyrode’s solution to 150 mL through a 14-channel Foley catheter and drained as required. This model has been previously validated [8]. Finally, the pelvic nerves (PNs) were identified (Fig. 1A, B).

Intravesical pressure transients were generated by PN stimulation with an isolated Grass SD9 stimulator (AstroNova, Elgin, IL, USA) via bipolar platinum-iridium cuff electrodes (miniature nerve cuff electrode, WPI, Hitchin, Herts, UK; Fig. 1C). The stimulus protocol was a 3-second train of 0.1-msec square waves (1–40 Hz) every 90 seconds. Before recordings, stimulation was applied to the electrode when placed about 1 cm from the PN to ensure absence of field stimulation to the nerve or bladder. The stimulator output voltage was varied for each bladder to obtain a threshold (V_thr, Fig. 1D, 10 V in this example), and recordings were obtained at twice-V_thr; this was checked at intervals and did not change by more than 10% during the experiments.

Pressure transients were generated by direct BW stimulation with the same stimulation protocol, via custom-made titanium mesh electrodes (6.25 cm²) with a 0.25-µm platinum-coating. An electrode was sutured to the bladder base (Fig. 1E) and a reference Ag/AgCl electrode to the dome.

At the end of the ex vivo experiments, a BW section was cut from the mid-dome, the mucosa was removed, and a detrusor strip (5-mm length; <1-mm diameter) was prepared to record isometric contractions by electrical field stimulation (EFS), using the same protocol [5]. The pressure and tension transducers were connected to a data acquisition system (PowerLab 16/35, WPI, Hitchin, Herts, UK) running LabChart v8 software.

Female rats were anaesthetised (urethane, 1.4 g/kg, intraperitoneal), the right femoral artery was cannulated to record blood pressure and heart rate, and the right femoral vein was used for fluid infusion; the trachea was cannulated to ensure airway patency. The rectal temperature was 37°C with a homeothermic blanket system (Harvard Apparatus, Holliston, MA, USA). Anaesthetic depth was assessed by monitoring paw withdrawal reflexes, blood pressure, and heart rate: if required, additional anaesthesia and fluids were administered. Through a midline laparotomy, the bladder was reflected to 1 side, the PN was identified and separated from the uterine wall, and a custom-made cuff electrode [10] was placed around the distal part. Intravesical pressure was monitored by a suprapubic 25-G needle connected to a pressure transducer, after filling to 1.5 mL. For preganglionic nerve stimulation, the same protocol was used (STMISOLA, Biopac Systems Inc., Santa Barbara, CA, USA), with threshold measurement and the lack of field stimulation (Fig. 1F) checked as above. At the end of the experiment the rat was euthanised with sodium pentobarbital (200 mg/kg, intraperitoneal) and the bladder was immediately removed. A detrusor strip, after removal of the mucosa, was prepared in the same manner as for the pig preparations to record isometric contractions with postganglionic nerve stimulation.

Tyrode’s solution contained (mM): NaCl, 118.0; NaHCO₃, 24.0; NaH₂PO₄, 1.8; KCl, 4.0; MgCl₂, 1.0; CaCl₂, 1.8; sodium pyruvate, 5.0; glucose 6.0, 5% CO₂-95% O₂, pH 7.4. Lidocaine-HCl was diluted in Tyrode’s solution to a final 2% (w/v) solution [11]; hexamethonium was dissolved in Tyrode’s solution to a final concentration of 300 µM [12]. Sildenafil was dissolved in dimethylsulphoxide (DMSO) as a 10mM stock solution and diluted to 20 µM in Tyrode’s solution. The chemicals and drugs were obtained from Sigma-Aldrich (Gillingham, Dorset, UK).

Data were recorded with a PowerLab 16/35 acquisition system and LabChart v8 software. Because pressure transients were superimposed on spontaneous contractions, the amplitude was measured with respect to the lowest pressure immediately after relaxation (Fig. 2A, B). The pressure (P) or tension (T)-frequency relations were parameterised as:

where P_max (T_max) is the maximum estimated value at high frequencies, f, and f_1/2 is the frequency required to attain P_max/2 (T_max/2); m is a constant. P-f (T-f) curves were also normalised to values for each curve at 40-Hz stimulation. Data sets are given as mean±standard error of the mean (SEM), with n equalling the number of bladders; tension values were normalised to the mean preparation cross-sectional area. Differences between data sets were tested with analysis of variance and the post hoc parametric t-test adjusted for multiple comparisons.

Pressure or tension transients in response to electrical stimulation of the PN (Fig. 2A), BW (Fig. 2B) and isolated detrusor strips (Fig. 2C) are shown. Mean (±SEM) frequency-response curves are shown to the right of the sample traces. Lidocaine completely abolished responses in all 3 preparations, indicating they were all nerve-mediated. The insets show force-frequency plots for the respective modes of stimulation, and the data were fitted by equation 1. In each preparation, 2 curves at 30 minutes apart were generated as time-controls. The respective P_max values for PN stimulation were 5.8±2.3 and 6.2±1.8 cm H₂O (n=7, P>0.05) and for BW stimulation were 6.2±1.8 and 6.7±1.1 cm H₂O (n=8, P>0.05). T_max values for EFS of isolated preparations were 2.0±0.7 and 2.1±0.3 mN/mg (n=6, P>0.05).

A nicotinic receptor antagonist, hexamethonium (300 µM), was used to confirm the target of stimulation in the 3 preparations. Contractions were completely abolished with PN stimulation (0%±0% control, n=7), but unaffected by EFS to isolated preparations (102.8% ±2.7% control, n =6): BW stimulation contractions were partially reduced (32.6% ±2.7% control, n=8) but the f_1/2 values were not significantly different. Thus, PN stimulation was preganglionic, EFS was postganglionic, and BW stimulation elicited pre- and postganglionic responses.

To test the main question of the study, the frequency-dependence of responses to the 3 modes of stimulation was calculated, as f_1/2 values. There was no significant difference in the f_1/2 values (f_1/2: the frequency for half-maximal response, equation 1) among the 3 modes of stimulation (Fig. 3A), suggesting that there was no frequency-dependent filtering by the intermediate ganglion. This result was consistent across the first and second sets of force-frequency curves from the time-control studies. In addition, the f_1/2 values of BW stimulation responses in the presence of hexamethonium (postganglionic responses) were not significantly different from those obtained from detrusor strips (7.8±1.7 Hz vs. 6.9±0.8 Hz, respectively, n=6, 6; P>0.05).

A second objective of the study was to identify whether sildenafil had any effect on frequency-dependent contractile responses. Force-frequency curves were generated in the presence of 20 µM sildenafil, and P_max or T_max was calculated. With postganglionic (EFS) stimulation, sildenafil reduced T_max by about 30% (Fig. 3B). However, the effect on preganglionic (PN) stimulation was much greater, with P_max reduced by about 75%. The effect on mixed pre- and postganglionic stimulation (BW) was intermediate, with a P_max reduction of about 45%. Control experiments with DMSO, used to dissolve sildenafil for stock solutions, had no effect on isolated detrusor preparations (T_max: 106.0%±3.7% control, n=6) or in PN-stimulated preparations (P_max: 101.4%±4.3% control, n=3).

Similar experiments were performed with PN stimulation of in vivo preparations (Fig. 4A) and EFS using in vitro (Fig. 4B) rat preparations. The control quantitative data for the 2 preparations were: P_max 14.3±3.1 cm H₂O (n=6) and T_max 1.8±0.4 mN/mg (n=6). The respective f_1/2 values were: 7.3±1.4 Hz (n=6) and 5.4±1.7 Hz (n=6), which were not significantly different from each other (P>0.05); the frequency-response curves are shown in Fig. 4C.

The main result of the study is that the frequency-dependence of contractions elicited by PN stimulation, direct stimulation of nerves in the whole BW, or in isolated detrusor preparations of the pig bladder was similar. Thus, preganglionic or postganglionic nerve stimulation generates the same frequency-dependent contractile response. There was a similar conclusion in rat preparations when in vivo and in vitro responses were compared. Moreover, the range of frequencies was similar in both species and to those recorded in other animals, as well as in human preparations [7,13,14]. All responses were nerve-mediated, inferred from their abolition at all frequencies by lidocaine; similar to that achieved by the neurotoxin tetrodotoxin [13]. This implies that action potential transmission between preand postganglionic fibres through the BW intramural ganglia is not frequency-modulated under these experimental conditions. Hypogastric sympathetic fibres do project to intramural ganglia, which could exert modulatory influences on intramural transmission, as exemplified in cat and dog preparations [13, 15]; however, this was not explored in this study. It was also of interest that pure postganglionic nerve stimulation in pig preparations, whether by EFS of isolated preparations or by BW stimulation in the presence of hexamethonium, produced the same frequency-dependent responses. This implies that dissection of muscle strips itself does not alter the characteristics of nerve-mediated responses.

The nicotine receptor antagonist, hexamethonium, completely abolished PN-mediated contractions in the ex vivo perfused pig bladder, which is interpreted as indicating that ACh is the sole transmitter acting on ganglionic nicotinic receptors. Similar contractions induced in anaesthetised rats were only partially abolished (not shown), and higher doses could not be used as the physiological status of the preparation would have been compromised. This limitation highlights an advantage of the ex vivo arterially-perfused bladder preparation, when higher doses of agents might be required. The partial abolition by hexamethonium of responses to direct pig BW stimulation implies that both pre- and postganglionic fibres are stimulated, corroborated by a lack of action on EFS responses in detrusor strips. Of interest, functional nicotinic receptors have been demonstrated in bladder urothelial cells and their modulation may contribute to the nicotine-induced increase of the cystometric intercontraction interval [16]. However, the experiments here using detrusor strips, denuded of mucosa, were not consistent with the presence of functional nicotinic receptors in the smooth muscle layer.

Sildenafil, a PDE5 inhibitor, increases intracellular cGMP and is associated with a relaxatory effect on isolated detrusor preparations. This may be due to a direct effect on detrusor contractile function [17], as well as inhibition of nerve-mediated release from postganglionic fibres, although ACh release remains intact [5]. The results presented here imply that sildenafil has actions at sites other than the BW, as the proportional reduction of contractile responses was significantly greater with preganglionic stimulation than with postganglionic stimulation. An interpretation is that sildenafil also affects preganglionic transmission. Several studies have demonstrated both stimulatory or inhibitory actions of the NO-cGMP pathway on transmitter release in central and peripheral pathways [18,19], implying there are multiple pathways that regulate release [20]. However, it is evident that agents such as PDE5 inhibitors that may be used as therapeutic agents on peripheral targets also have central actions; with respect to the lower urinary tract, this would further depress bladder function.

These data are useful to identify a frequency range for bladder stimulation using neuromodulation devices, where too low a frequency would achieve inadequate stimulation and too high a frequency would reduce stimulator efficiency and potentially tissue damage. Sacral nerve stimulators have pulse rates of about 1–150 Hz, with individual pulses of 50–500 µsec [21], which may be compared to the 100-µsec pulses at 1–40 Hz in the current study. A short pulse width (<50–100 µsec) is important, as this can selectively stimulate nerve fibres without direct action on smooth muscle, due to the different membrane time constants of these 2 cell types. Recent studies of rechargeable sacral nerve stimulators have shown success and patient satisfaction with devices delivering 200-µs pulses at about 14 Hz [22]. More recently, intravesical devices are also being trialled and a recent study [23] to manage OAB contractions used 200-µsec pulses at frequencies between 5 and 50 Hz. The use of isolated detrusor preparations is an efficient way to characterise the properties of nerve-mediated contractions. The study shows that the frequency-dependent properties of isolated preparations can be used to set appropriate neurostimulation variables applied to the PN or BW.