Characterisation of nerve‐mediated ATP release from bladder detrusor muscle and its pathological implications

Abstract

Background and Purpose

This study aims to characterise the molecular mechanisms that determine variability of atropine resistance of nerve‐mediated contractions in human and guinea pig detrusor smooth muscle.

Experimental Approach

Atropine resistance of nerve‐mediated contractions and the role of P2X1 receptors, were assessed in isolated preparations from guinea pigs and also humans with or without overactive bladder syndrome, from which the mucosa was removed. Nerve‐mediated ATP release was measured directly with amperometric ATP‐sensitive electrodes. Ecto‐ATPase activity of guinea pig and human detrusor samples was measured in vitro by measuring the concentration‐dependent rate of ATP breakdown. The transcription of ecto‐ATPase subtypes in human samples was measured by qPCR.

Key Results

Atropine resistance was greatest in guinea pig detrusor, absent in human tissue from normally functioning bladders, and intermediate in human overactive bladder. Greater atropine resistance correlated with reduction of contractions by the ATP‐diphosphohydrolase apyrase, directly implicating ATP in their generation. E‐NTPDase‐1 was the most abundantly transcribed ecto‐ATPase of those tested, and transcription was reduced in tissue from human overactive, compared to normal, bladders. E‐NTPDase‐1 enzymic activity was inversely related to the magnitude of atropine resistance. Nerve‐mediated ATP release was continually measured and varied with stimulation frequency over the range of 1–16 Hz.

Conclusion and Implications

Atropine resistance in nerve‐mediated detrusor contractions is due to ATP release and its magnitude is inversely related to E‐NTPDase‐1 activity. ATP is released under different stimulation conditions compared with ACh, implying different routes for their release.

What is already known

• Atropine resistance in detrusor results from nerve‐mediated ATP release.

• Atropine resistance when it occurs in human detrusor is from patients with overactive bladder pathologies.

What this study adds

• Atropine resistance results from incomplete hydrolysis of ATP at the nerve–muscle junction.

• Nerve‐mediated ATP release occurs at a lower stimulation frequency range that generate contractions.

What is the clinical significance

• With human samples, functional nerve‐mediated ATP release is associated with overactive bladder.

• Selective modulation of nerve‐mediated ATP offers a drug target to manage overactive bladder.

Abbreviations

ABMA

α,β‐methylene ATP

ARL‐67156

6‐N,N‐diethyl‐d‐β,γ‐dibromomethylene ATP

EFS

electrical field stimulation

E‐NTPDase

ecto‐nucleoside triphosphate diphosphohydrolase

human DO

tissue from patients with detrusor overactivity

human stable

tissue from patients with normal bladder function

IDO

idiopathic detrusor overactivity

NDO

neuropathic detrusor overactivity

TTX

tetrodotoxin

1. INTRODUCTION

Contraction of urinary bladder detrusor smooth muscle is initiated by excitation of postganglionic parasympathetic fibres that release https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=294 and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1713. With human detrusor muscle from normal bladders, ACh is the sole functional transmitter as atropine abolishes nerve‐mediated contractions (Bayliss, Wu, Newgreen, Mundy, & Fry, 1999). However, with detrusor from most other mammals, part of the contraction is resistant to the action of https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=320. Atropine‐resistant contractions are proposed to be mediated by ATP acting on https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=478 receptors (Lee, Bardini, & Burnstock, 2000) as they are greatly attenuated by the non‐hydrolysable analogue of ATP, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4093: Palea, Artibani, Ostardo, Trist, & Pietra, 1993; Peterson & Noronha‐Blob, 1989) by rapid desensitisation of the receptor (North & Surprenant, 2000). In addition, ATP release is associated with nerve‐mediated detrusor contractions (Burnstock, Cocks, Crowe, & Kasakov, 1978; Hashitani & Suzuki, 1995). However, some studies have proposed that ATP also activates other P2X receptors, in particular, a P2X1/4 receptor heteromer (Kennedy, Tasker, Gallacher, & Westfall, 2007; Syed & Kennedy, 2012).

With human detrusor, atropine resistance of nerve‐mediated contractions occurs in particular with advancing age (Yoshida et al., 2001; but see Yokota & Yamaguchi, 1996) and with overactive bladder symptoms accompanying several pathologies, including neurological injuries, outflow tract obstruction and idiopathic causes (Bayliss et al., 1999). Similar pathologies in animal models also increase the proportion of the purinergic component of nerve‐mediated contractions (Moss, Tansey, & Burnstock, 1989; Mumtaz et al., 2006).

One area of this study was to determine why atropine resistance varies between species (human and guinea pig) and why it occurs more in functional bladder pathologies. Several hypotheses may be proposed for this variability: (a) the potency and efficacy of P2X1 receptor agonists to generate responses may vary with detrusor from different species and pathologies; (b) detrusor muscle not exhibiting atropine‐resistant nerve‐mediated contractions may not release ATP; and (c) nerve‐mediated ATP release always occurs but is variably hydrolysed in the nerve‐muscle junction by ecto‐ATPases and in some detrusor preparations may not activate the detrusor muscle. A nucleotide‐specific group of ecto‐ATPases is the https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2888 (E‐NTPDases), which have eight paralogues (Zimmermann, Zebisch, & Sträter, 2012).

Further insight into how ATP‐dependent contractions may be selectively manipulated comes from evidence that ATP and ACh may be released from postganglionic nerve terminals by different pathways. Indirect evidence suggests that ATP is released at lower stimulation rates, compared to those releasing ACh (Calvert et al., 2001; Chakrabarty et al., 2019; Pakzad et al., 2016). There is more direct evidence that the https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1304&familyId=260&familyType=ENZYME inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4743 abolishes nerve‐mediated ATP release (Chakrabarty et al., 2019); however, this should be set against the observation that ACh is also modulated by agents such as https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2844 (Silva‐Ramos et al., 2015). The development of amperometric ATP‐selective electrodes potentially allows for the real‐time measurement of nerve‐mediated ATP release in detrusor muscle. Measurement of the frequency dependence of ATP was attempted in his study to relate it to the contractile data that suggests frequency dependence of release of different neurotransmitters.

The aim of this study was to elucidate how atropine‐resistant, ATP‐dependent contractions are generated. As these are a feature of detrusor overactivity in the human bladder, this should provide targeted drug models to attenuate specifically this particular bladder pathology.

2. METHODS

2.1. Ethical approval and tissue sources

All procedures involving human tissues were in accordance with the approval of the ethical committee at University College London Hospitals and the 1964 Helsinki Declaration. Human and guinea pig bladder detrusor muscle was used. Human biopsies were obtained at open surgery from patients with idiopathic (n = 5; 50 ± 15 years; idiopathic detrusor overactivity [IDO]) or neuropathic (n = 6; 34 ± 10 years; neuropathic detrusor overactivity [NDO]) detrusor overactivity (human DO) or those undergoing cystectomy with no DO symptoms (human stable, n = 9; 57 ± 14 years). NDO and IDO data were not significantly different in any variable and were merged. Patient ages of the merged DO and the stable groups were not statistically different. Biopsies were brought to the laboratory in ice‐cold Ca²⁺‐free Tyrode's solution within 1 hr of excision and used immediately.

All animal care and experimental procedures were in compliance with the University of Bristol Ethics Committee (approval 17.09.2014) and carried out in accordance with the UK Animals (Scientific Procedures) Act 1986. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2010; McGrath & Lilley, 2015) and with the recommendations made by the British Journal of Pharmacology. Adult guinea pigs were used as detrusor function has been previously well characterised. Animals (Dunkin‐Hartley, males, 350–400 g) were obtained from the Animal Services Unit, University of Bristol, housed singly in straw‐floored cages at 22°C with a 12‐hr light–dark cycle and with water and food available ad libitum. Animals were killed by a Schedule 1 procedure; by injection with Na https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5480 (200 mg·kg⁻¹, i.p.); and by cervical dislocation, verified by a lack of corneal and spinal reflexes, and the bladder was immediately removed through a laparotomy.

The mucosa (urothelium and lamina propria) was removed from human and animal tissue, and detrusor strips (<1‐mm diameter, 5‐mm length) were dissected, in Ca²⁺‐free Tyrode's solution, for tension and ATP release experiments. The remainder of the tissue was cut into three or four pieces (≈20 mg each) with a fresh sharp razor blade and frozen in liquid‐N₂ for RNA extraction and ATPase activity measurement.

2.2. Tension recording, nerve‐mediated ATP‐release, and measurement of intracellular Ca²⁺

Detrusor strips were tied to an isometric force transducer and a fixed anchor in a horizontal trough and superfused with Tyrode's solution at 4 ml·min⁻¹. Electrical field stimulation (EFS), via Pt plates in the sides of the trough, used 0.1‐ms pulses in 3‐s trains (frequency of 1–32 Hz) every 90 s. Concentration–response curves for ATP and ABMA were constructed with unstimulated preparations, using test concentrations between 10⁻⁶ to 2.10⁻² M for ATP and 10⁻⁸ to 10⁻⁴ M for ABMA in equal half‐log increments. Force–frequency or concentration–response curves were fitted to Equation (1):

$$T=\left(T_{\mathit{max}}{x}^n\right)/\left(x^n+x_{1/2}^n\right).$$ (1)

T _max is the maximum response at high frequencies or concentrations, x; x _1/2 the frequency (f _1/2) or concentration (EC₅₀) required to elicit T_max/2; and n is a constant. To measure the effect of ABMA (1 μM) on nerve‐mediated contractions, EFS was stopped after control recordings; ABMA was added to the superfusate and after relaxation of the resulting contracture EFS recommenced (about 15 min); and ABMA was then washed out. For atropine (1 μM) or https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2888 (10 U·ml⁻¹), the agent was added to the superfusate, with the preparation stimulated throughout at 8 Hz.

Amperometric ATP electrodes (Sarissa Biomedical Ltd, Coventry, UK) were used to measure nerve‐mediated ATP release with the active tip (2‐mm length, 50‐μm diameter) placed on the surface of the preparation parallel to the longitudinal axis. A null electrode, lacking the sensing surface, was similarly placed and both polarised to 0.65 V. Electrode outputs formed the inputs to a home‐made differential amplifier with high common mode rejection, and the output was recorded to attenuate stimulation artefacts. The Tyrode superfusate contained 2‐mM glycerol, required for the enzymatic detection of ATP. Prior to experiments, the system was calibrated by exposure to 10‐μM Na₂ATP—electrodes had a linear response between 0‐ and 10‐μM ATP (see Figure 1a). ATP transients were elicited by EFS (1–24 Hz). Two ATP/force–frequency relationships at 20‐min intervals were done as time controls of the percentage of second compared to the first calculated. At 8 Hz: tension 101.4 ± 10.6%; ATP 100.2 ± 15.3%, at 12 Hz: tension 102.2 ± 11.9%; ATP 98.8 ± 11.6% (n = 5).

Figure 1.

ATP transients in guinea pig detrusor muscle preparations. (a) Recordings of isometric tension (upper blue trace) and outputs from ATP‐selective electrode and null electrodes (lower black traces), the difference recording (ATP–null, red trace) is also shown. Stimulation was a 3‐s train at 2 Hz. Arrows above the tension and ATP–null traces show the respective times of peak values. The inset shows a calibration curve of an ATP electrode with a linear fit, as well as a sample calibration trace. (b) Tension (blue) and ATP–null (red) traces in the presence of 10‐μM carbachol. (c) Tension (lower) and ATP–null (upper) traces in the absence and presence of 1‐μM tetrodotoxin (TTX), in this example at 8‐Hz stimulation. (d) Tension (lower) and ATP–null (upper) traces in the absence and presence of 1‐μM atropine, in this example at 4‐Hz stimulation

In a separate series of experiments to measure nerve‐mediated release at a fixed frequency (8 Hz), superfusate samples (100 μl) were taken at a fixed distance (1 mm) above the preparation and 2 s after initiation of stimulation and analysed by a luciferin‐luciferase assay as described previously (Kushida & Fry, 2016).

Myocytes were isolated from detrusor strips by collagenase dispersion. Fura‐2 AM (5 μM) was added to cell suspensions for recording intracellular [Ca²⁺] during control periods on an exposure to ABMA. The cell preparation procedure, experimental protocol, and signal calibration have been explained in detail (Montgomery & Fry, 1992; Wu & Fry, 2001); all experiments were at 36°C.

2.3. Gene expression of E‐NTPDases

Total RNA was extracted from frozen tissue (30 mg) using an RNeasy Mini Kit (Qiagen, UK) as per manufacturer's instructions. RNA integrity was determined with an Aligent 2100 bioanalyser using the 18S and 28S ribosomal RNA bands as controls. This clearly showed visible single peaks indicative of high‐quality RNA: The RNA concentration of each sample (7 μl) was determined with a Genequant 1300 spectrophotometer (VWR, UK). cDNA was synthesised from each RNA sample and then used for qPCR reactions using specific primers for E‐NTPDase‐1https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2889&familyId=799&familyType=ENZYME ‐3, and ‐5. E‐NTPdase‐1, ‐2, and ‐3 were chosen as they are extracellular enzymes, E‐NTPDase‐5, although intracellular, may be secreted (Zimmermann et al., 2012). The resulting amplified RT‐PCR products (TaqMan system, ThermoFisher Scientific) were separated by 1.5% agarose gel electrophoresis and visualised with SyberGold (Molecular Probes), quantified and expressed as a proportion of 18S cDNA.

2.4. Measurement of ecto‐ATPase activity

Frozen detrusor samples (four 20‐mg samples per bladder or biopsy sample) were separately thawed in 3‐ml Ca²⁺‐free HEPES Tyrode's and then equilibrated at 37°C in 3‐ml Tyrode's for 30 min. Samples were then transferred to 980‐μl Tyrode's, two of which contained the E‐NTPDase inhibitor 6‐N,N‐diethyl‐d‐β,γ‐dibromomethylene ATP (https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=9030; 100 μM). After a further 10 min, 10‐mM Na₂ATP stock (20 μl in Tyrode's) was added for a final [ATP] of 0.2 mM. Subsequently, 10‐μl aliquots were added to 1.99‐ml Ca²⁺‐free HEPES Tyrode's with 5‐mM EDTA at 0, 5, 10, 20, and 30 min for ATP analysis by a luciferin‐luciferase assay (GloMax 20/20, Promega, UK): The initial rate of ATP breakdown was calculated. After 30 min, the samples were washed in 3.5‐ml Tyrode's and the ATP breakdown rate at 0.5‐mM initial [ATP] re‐commenced, the cycle was repeated for initial [ATP] of 1.0, 2.0, and 5.0 mM. A final run in 0.2‐mM initial [ATP] was done, the initial rate was compared to the first estimate and used if within 10%. Finally, tissue samples were weighed and assayed for protein content (Bradford Assay, ThermoFisher Scientific). ATP concentrations at all time and concentration points in the presence of ARL‐67156 were subtracted from those in its absence and used as the ENDTPase (ARL‐dependent) rate. Initial rate, V _ATP, was plotted as a function of starting [ATP] to estimate the maximum rate at high concentrations, V _max, and the K _M of the reaction: V = (V_max[ATP])/(k_m+[ATP]).

2.5. Data presentation and analyses, experimental design

The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology (Curtis et al., 2018). Most data are means ± SD (n = number of separate human biopsies or guinea pig bladders). Data for the effects of atropine and apyrase are expressed as medians [25,75% interquartiles] as they were highly skewed sets in some instances. Data sets were compared by ANOVA followed by post hoc Fisher's least significant difference comparison only if F was significant and there was no variance inhomogeneity: A value of P < .05 (^*) was accepted as significantly different. No data outliers were excluded. KaleidaGraph (RRID:SCR_014980) was used for data analysis and curve fitting with a non‐linear iterative fit programme. Sample size calculations (http://www.3rs-reduction.co.uk) used previous experimental data (Harvey, Skennerton, Newgreen, & Fry, 2002; Pakzad et al., 2016) with animal and human tissue suggested group sizes of n = 5–6 for 80% power and .05 for a Type‐I error. Individual data points to compile summaries in Table 1 and Figure 3 are shown in Figure S1a–i). Experiments were either interventional or compared data from either normal or pathological human bladder samples, and no randomisation or blinding of samples was undertaken.

Table 1.

Values of ATP‐dependent nerve‐mediated force of contraction, responses to ABMA, ecto‐ATPase activity, and E‐NTPDase‐1 expression

Variable	Human stable	Human overactive	Guinea pig
Nerve‐mediated contractions
Control, mN·mm−2 (8 Hz stim)	7.9 [6.6, 9.5] (19)	6.1 [3.6, 9,2] (16)	10.3 [8.5, 11.5] (7)
+ atropine	0.0 [0.0, 0.08] (19)	0.6 [0.4, 1.2] (16)*	3.2 [2.8, 3.8] (7)*, §
Atropine resistance, % total	0.0 [0.0, 1.0] (19)	9.1 [5.7, 29.9] (16)*	31.2 [26.6, 43.1] (7)*, §
Control, mN·mm−2 (8 Hz stim)	8.5 ± 2.6 (14)	11.2 ± 4.7 (11)	11.4 ± 3.3 (11)
+ ABMA	7.4 ± 2.6 (14)	7.3 ± 4.4 (11)#	8.9 ± 2.9 (11)#
Apyrase reduction, % control	1.7 [−1.1, 4.8] (7)	11.1 [5.7, 12.0] (9)*	31.9 [23.6, 42.8] (9)*, §
ABMA, ATP potency, and ABMA efficacy
ABMA pEC50, strips	5.51 ± 0.11 (7)	5.41 ± 0.12 (7)	5.53 ± 0.13 (7)
ABMA pEC50, myocytes	6.77 ± 0.23 (7)	6.65 ± 0.10 (7)	6.54 ± 0.18 (7)
ABMA efficacy mN·mm−2, strips	13.4 ± 5.6 (15)	15.2 ± 4.3 (15)	12.1 ± 3.1 (9)
ATP, pEC50	3.11 ± 0.77 (20)	3.73 ± 0.91 (16)*	3.86 ± 0.40 (8)*
Ecto‐ATPase activity
V max‐total, nmol·mg−1·s−1	1.89 ± 0.68 (8)	1.15 ± 0.34 (11)	1.19 ± 0.55 (7)
K M‐total, mM	1.46 ± 0.30 (8)	1.38 ± 0.26 (11)	1.07 ± 0.15 (7)
V max (ARL‐sens), nmol·mg−1·s−1	0.98 ± 0.25 (8)	0.60 ± 0.16 (11)*	0.37 ± 0.07 (7)*, §
K M (ARL‐sens), mM	1.38 ± 0.47 (8)	1.57 ± 0.46 (11)	1.35 ± 0.49 (7)
E‐NTPDase transcription
E‐NTPDase‐1/18S.10−4	3.88 ± 1.28 (9)	2.61 ± 1.13 (9)*
E‐NTPDase‐2/18S.10−4	0.035 ± 0.023 (9)	0.029 ± 0.027 (9)
E‐NTPDase‐3/18S.10−4	0.090 ± 0.070 (9)	0.047 ± 0.050 (9)
E‐NTPDase‐5/18S.10−4	0.11 ± 0.05 (9)	0.14 ± 0.12 (9)

Note. Number of preparations in parentheses. Data are mean ± SD, except for the atropine‐resistance and apyrase‐reduction data which are median [25,75% interquartiles] due to the skewed nature of some of these data sets. See Figures S1–S8 for individual data points used to compile Table 1.

^* P < .05, significantly different from human stable.

^§ P < .05, significantly different from human overactive.

^# P < .05, ABMA significantly different from control.

Figure 3.

ATP‐transients and nerve‐mediated contractions. Data from guinea pig preparations. (a) Frequency dependence of tension (lower, blue traces) and ATP–null (upper, red traces) traces. (b) Dependence of the peak ATP transient (closed, red circles) and tension (black, closed squares) on stimulation frequency. Values of the half‐maximal frequency (f _1/2) for tension (T, f _1/2,T) are shown. The mean value for the ATP‐transient magnitude at 24 Hz was not used for the curve fit. Also shown is the frequency dependence of tension in the presence of 1‐μM atropine (closed, blue squares). Data are mean ± SD, n = 18 for tension values and n = 6 for ATP data. See Figure S1i for f _1/2,T and f _1/2,ATP values in individual preparations

2.6. Materials

Tension and ATP release experiments were at 37°C in Tyrode's solution (mM): NaCl, 118; KCl, 4.0; NaHCO₃, 24; NaH₂PO₄, 0.4; MgCl₂, 1.0; CaCl₂, 1.8; glucose, 6.1; pyruvate, 5.0; gassed with 95% O₂, 5% CO₂ (pH 7.45 ± 0.03). Ca²⁺‐free Tyrode's solution contained HEPES (10 mM) + NaCl (14 mM) to replace NaHCO₃, pH 7.4 with 1‐M NaOH and gassed with 100% O₂. Atropine, apyrase, ABMA, https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2616 (TTX), ARL‐67156, and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=298 were stored as aqueous stocks and added to Tyrode's for appropriate final concentrations. All chemicals were from Sigma, UK.

2.7. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander, Christopoulos, et al., 2017; Alexander, Fabbro, et al., 2017; Alexander, Peters, et al., 2017).

3. RESULTS

3.1. ATP dependence of nerve‐mediated contractions

EFS contractions in human and guinea pig preparations were abolished by TTX (1 μM); in addition, contractions from human‐stable bladders were also completely abolished by atropine (1 μM). However, with preparations from overactive human bladders (human DO) and guinea pigs, atropine did not completely block contractions, leaving an atropine‐resistant component: The absolute tension values and the percentage of the contraction remaining are shown in Table 1 (and Figure S1a,b). Involvement of ATP as a neurotransmitter was indicated in four ways: (a) ABMA (1 μM) abolished atropine‐resistant contractions (not shown: see also Bayliss et al., 1999; Peterson & Noronha‐Blob, 1989); (b) the ecto‐ATPase inhibitor ARL‐67156 increased guinea pig detrusor contractions to 1.42 ± 0.15 times control (n = 5) and reduced f _1/2 values (5.5 ± 0.7 to 4.8 ± 0.9 Hz, n = 5); (c) pretreatment with ABMA reduced significantly contractions from human‐DO and guinea pig detrusor but not from human‐stable bladders (Table 1 and Figure S1c); and (d) the non‐specific ATPase, apyrase had no effect on detrusor contractions of human‐stable bladder but reduced those from human‐DO and guinea pig bladders (Table 1 and Figure S1d). Moreover, the percentage reduction by apyrase was similar to the atropine‐resistant percentage in the three groups.

The reason for variable atropine resistance in the three cohorts may be explained by the following: (a) the potency and efficacy of agonists at detrusor P2X1 receptors are different; (b) the amount of ATP released by motor nerves varies; and (c) ATP is hydrolysed to varying extents at the neuromuscular junction. These possibilities were subsequently tested.

3.2. Potency and efficacy of ABMA

Contractions from detrusor strips in the absence of EFS and intracellular Ca²⁺ transients from isolated myocytes generated by addition of ABMA to the superfusate were recorded. With detrusor strips and isolated myocytes, the pEC₅₀ values were not significantly different between the three groups (Table 1and Figure S1e). The efficacy of 1‐μM ABMA to generate tension was also similar between the groups (Table 1 and Figure S1e). Thus, variability of atropine resistance between the three groups cannot be due to differences of detrusor responsiveness to ABMA. It has been reported that ATP may act via receptors in addition to the P2X1 receptors (Kennedy et al., 2007), the latter desensitised by ABMA. This was tested in six human samples (three with idiopathic DO and three normal) by adding ATP (1 mM) before and after exposure to ABMA. The response after ABMA was 6.5 ± 2.5% of that before (with no difference in the values obtained with normal or DO samples) and suggests that a small fraction of the response to ATP is via a receptor other than P2X1 receptors.

3.3. Nerve‐mediated release of ATP

Real‐time ATP release was measured with amperometric ATP electrodes during EFS (1–24 Hz for all interventions) from guinea pig preparations: Figure 1a shows an example of ATP‐electrode and tension recordings from a preparation stimulated, in this case, at 2 Hz. Electrode responses are from ATP and null electrodes, with the differential recording (ATP–null) used for analysis. The arrows at the peak of the tension and ATP–null traces mark the time to maximum response, where the tension peak always preceded the ATP–null response (see Section 4). The inset of Figure 1a shows an example of an ATP calibration pulse with a linear calibration curve constructed for several ATP concentrations. ATP‐electrode responses were unaffected by the muscle contraction itself as no response was elicited with 10‐μM carbachol added to the superfusate (n = 7; Figure 1b).

EFS‐induced contractions and ATP transients were abolished by 1‐μM TTX (n = 9, Figure 1c, 8 Hz in this example). ATP transients were unaffected by 1‐μM atropine in magnitude or duration (11.1 ± 1.4 and 10.3 ± 2.1 s, without or with atropine, n = 6), although contractions were reduced (n = 6, Figure 1d, 4 Hz in this example). Nerve‐mediated ATP release, in the presence of ARL‐67156 to reduce ATP breakdown, was also measured in the three cohorts (guinea pig; human stable; and human DO) at a fixed frequency (8 Hz) using a luciferin‐luciferase assay: Values were not significantly different from each other: 0.81 ± 0.41; 1.03 ± 0.57; and 0.62 ± 0.30 pmol·μl⁻¹ (all n = 5).

3.4. ecto‐ATPase activity and E‐NTPDase expression

In contrast to ABMA, ATP exhibited a variable potency on detrusor contractions in the different cohorts: The ATP pEC₅₀ was lower in human‐stable preparations than in human‐DO and guinea pig preparations (Table 1 and Figure S1f). This may be due to a differential hydrolysis of ATP in the nerve–muscle junction. This was tested by measuring ecto‐ATPase activity in human and guinea pig detrusor tissue, as well as gene expression for extracellular ATPases (E‐NTPDases) in human detrusor from stable and overactive bladders.

Detrusor samples showed ATPase activity that was partly reduced by ARL‐67156 (Figure 2a, sample experiment from human‐stable preparation). Reactions were analysed by calculation of V _max and K _M values for total ATPase activity, as well as the ARL‐67156‐dependent fraction: the latter was used as an estimate of ecto‐ATPase activity. The V _max for the ARL‐67156‐dependent component (V _max‐ARL) was significantly greater in detrusor from human‐stable bladders, compared to those from human‐DO bladder and from guinea pig bladder; these latter were not significantly different (Table 1 and Figure S1g). K _M values for total (K _M‐total) and ARL‐dependent (K _M‐ARL) ATPase activities were similar for all three groups. Furthermore, V _max‐ARL and K _M‐ARL values for detrusor from idiopathic (IDO, n = 5) and neurogenic (NDO, n = 6) were similar; the merged data set is shown in Table 1 (IDO vs. NDO: V _max‐ARL: 0.66 ± 0.19 vs. 0.55 ± 0.12 nmol·mg prot⁻¹·s⁻¹; K _M‐ARL: 1.63 ± 0.25 vs. 1.52 ± 0.79 mM,).

Figure 2.

ecto‐ATPase activity in detrusor smooth muscle. (a) Sample experiment from human‐stable detrusor tissue of the initial rate of ATP hydrolysis as a function of the starting ATP concentration, in the absence (total) and presence (ARL independent) of 100‐μM ARL‐67156. The difference between the two (ARL dependent) is also plotted as a measure of ecto‐ATPase activity. The V _max and K _M values of the ARL‐dependent fraction are shown. (b) The relationship between ecto‐ATPase V _max and the percentage purinergic component of the contraction (8‐Hz stimulation) as determined by (a) the percentage residual contraction with atropine (closed circles) or percentage reduction of the contraction by apyrase (open circles). Numbers of preparations in the three cohorts of tissue, contributing to the contractile data (ordinate) and ecto‐ATPase data (abscissa), are shown in Table 1 and Figure S1b,d,g

The dependence of percentage atropine resistance or reduction by apyrase on V _ATP,max‐ARL is shown in Figure 2b. Thus, the human‐stable bladder samples had no ATP‐dependent component of the nerve‐mediated contraction (no atropine resistance or reduction by apyrase), and the highest V _max‐ARL value, whereas the guinea pig detrusor samples were at the other end of the range, with human‐DO detrusor samples in an intermediate position.

RNA expression of four ecto‐ATPase subtypes (E‐NTPDase‐1, ‐2, ‐3, and ‐5) was measured in tissue from human‐stable and human‐DO bladders. E‐NTPDase‐1, ‐2, and ‐3 are extracellular enzymes, E‐NTPDase‐5 whilst intracellular may be secreted and so was included (Robson, Sévigny, & Zimmermann, 2006; Zimmermann et al., 2012). E‐NTPDase‐1 was expressed most in human tissue, but expression was significantly less in tissue from overactive bladders (Table 1 and Figure S1h), consistent with reduced ecto‐ATPase activity. Consistent with variable ATPase activity in the three cohorts is that ATP dose–response curves for contracture development showed different pEC₅₀ values (Table 1). Thus, the pEC₅₀ was lowest in human‐stable detrusor, greatest in guinea pig tissue, and intermediate in human‐DO tissue.

3.5. Frequency dependence of nerve‐mediated ATP release

ATP transients were recorded in guinea pig detrusor by EFS using 3‐s trains of stimuli at frequencies from 1 to 24 Hz (Figure 3a). The magnitude of the ATP transients increased with stimulation frequency but reached a maximum at lower frequencies than did contractions. Thus, tension and ATP transients showed different f _1/2 values (f _1/2 = frequency for half‐maximal response, Figure 3b); tension 7.8 ± 2.0 Hz (n = 18) versus ATP 2.2 ± 1.0 Hz (n = 6). At higher frequencies, the contour of the ATP transient also changed in some experiments, with a distinctive tail and the peak amplitude sometimes diminishing. Thus, the integral of the ATP transient over 10 s (∫ATP₁₀) was also calculated and the f _1/2 values were again estimated: f _1/2 values for the ATP amplitude and ∫ATP₁₀ were not significantly different (2.2 ± 1.0 vs. 3.4 ± 0.8 Hz, n = 6). In the presence of atropine, the force–frequency curve shifted to the left and was similar to the ATP–frequency curve (f _1/2 values: 3.6 ± 1.4 Hz vs. 2.2 ± 1.0 Hz, —individual data values in Figure S1i). This is consistent with the hypothesis that the tension–frequency curve in the presence of atropine is determined by nerve‐mediated ATP release.

4. DISCUSSION

4.1. Atropine‐resistance and nerve‐mediated ATP release

Atropine resistance of nerve‐mediated detrusor contractions is a well‐established phenomenon in bladder tissue from most small animals and pathological human bladders, but absent in the human‐stable bladder. There is substantial, albeit mainly indirect, evidence that atropine resistance results from nerve‐mediated release of ATP, in addition to the normal secretion of ACh. This is interpreted from the abolition of atropine‐resistant contractions by ABMA, an agent which desensitises P2X1 receptors (Bayliss et al., 1999; Palea et al., 1993). However, it must be appreciated that ABMA is also an agonist at other P2X receptor subtypes (Lê, Babinski, & Seguela, 1998). ATP‐dependent contractions are more rapid than those mediated by ACh, and in animals, rapid, partial urine voids are used as territorial marking (Desjardins, Maruniak, & Bronson, 1973). However, there is equal interest regarding their presence in human storage and voiding pathologies and they may be responsible for overactive or non‐voiding contractions. Apyrase is a highly active ATP‐diphosphohydrolase and reduced nerve‐mediated contractions by almost the same proportion that residual contractions were recorded in the presence of atropine. Thus, apyrase and atropine may reveal the same fraction of the contraction, namely, that mediated by ATP. Some reports showed a small (6.5%) purinergic component resistant to P2X1 receptor antagonists, assuming that ABMA primarily desensitises these receptors (Kennedy, 2015; Kennedy et al., 2007).

4.2. Causes of atropine resistance

Several hypotheses were tested to account for the variable appearance of ATP‐dependent nerve‐mediated contraction in three cohorts: (a) the potency and efficacy of detrusor for P2X receptors were different; (b) there was variable ATP release by the motor nerve; and (c) ATP was hydrolysed to different extents in the neuromuscular junction. The potency and efficacy of detrusor for ATP was tested in muscle strips and isolated myocytes from guinea pig, human‐stable, and human‐DO bladders, with no significant differences in muscle strips or in isolated cells. This functional observation in human detrusor is consistent with unchanged expression of P2X1 receptors (O'Reilly et al., 2002) in detrusor from human pathological bladders, except an increase in obstructed bladders (O'Reilly et al., 2001). However, in rat tissue superfused in an organ bath for several hours, there was a markedly reduced expression of P2X1 receptors (Elliott et al., 2013), although in this study, ABMA responses were stable over the experimental time course. Another possibility is that ATP is not released from nerves in human‐stable detrusor and is greatest in guinea pig tissue. However, nerve‐mediated ATP release was seen in all groups when ecto‐ATPase activity was attenuated. Even though actual values may be attenuated by any remaining endogenous ecto‐ATPase activity, it might argue against this being the principal cause for the lack of atropine‐resistant contractions in human‐stable detrusor.

The final possibility examined is that variable ecto‐ATPase activity and expression of ATPases themselves account for the different extent of atropine‐resistant contractions. Ecto‐ATPase activity was inversely associated with the magnitude of nerve‐mediated purinergic contractions and this was corroborated in human tissue by reduced ecto‐ATPase expression of the predominant enzyme E‐NTPDase‐1 in detrusor from overactive compared to stable bladders. E‐NTPDase‐1 is inhibited by ARL‐67156 (Lévesque, Lavoie, Lecka, Bigonnesse, & Sévigny, 2007) which would account for the increase of nerve‐mediated contractions in guinea pig detrusor (Westfall, Kennedy, & Sneddon, 1997) and validated its use to estimate ecto‐ATPase activity in detrusor homogenates.

Variation of ecto‐ATPase activity may also have an indirect effect through changing local concentrations of the products of ATP hydrolysis. E‐NTPDase‐1 is the key subtype in human detrusor (Silva‐Ramos et al., 2015) which metabolises ATP to https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2455, with further degradation to https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2844, but without intermediate accumulation of https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1712. Thus, the absence of atropine‐resistant contractions in human‐stable detrusor may be contributed by greater degradation of nerve‐mediated ATP release and also by adenosine itself attenuating ATP release via an https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=18 mechanism (Pakzad et al., 2016). Adenosine reduced nerve‐mediated contractions mostly at low frequencies, whilst ARL‐67156 increased contractions over the same frequency range making it difficult to distinguish between the two possibilities. Several arguments may favour variability of ecto‐ATPase activity as an explanation for variable atropine resistance: firstly, nerve‐mediated ATP release was similar in the three cohorts used in this study although atropine resistance varied greatly and secondly, the A₁ receptor agonist https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=380 had no effect on human‐stable detrusor but a significant action on guinea pig tissue (Pakzad et al., 2016) suggesting that adenosine‐mediated suppression of ATP, via A₁ receptors , was not a feature of human‐stable detrusor.

4.3. Frequency dependence of ATP release

Nerve‐mediated detrusor contractions showing atropine resistance have a component dependent on ATP, dominant at low frequencies, and another on ACh at higher frequencies and agree with previous observations (Brading & Williams, 1990; Pakzad et al., 2016; Werner, Knorn, Meredith, Aldrich, & Nelson, 2007). Here, we showed that direct measurement of nerve‐mediated ATP release was over the same range of frequencies that generated purinergic contractions, in the presence of atropine. The differential frequency dependence of ATP and ACh release conforms to the different cellular pathways that each transmitter regulates to generate contraction; a rapid P2X‐dependent activation of myosin light chain kinase by Ca²⁺–calmodulin and a slower muscarinic Ca²⁺ desensitisation of contractile proteins (Tsai, Kamm, & Stull, 2012). A similar frequency‐dependent release of ATP and noradrenaline is present in ear artery and vas deferens preparations innervated by sympathetic nerves, with ATP release at lower frequencies (Kennedy, Saville, & Burnstock, 1986; Todorov, Mihaylova‐Todorova, Craviso, Bjur, & Westfall, 1996). The peak of the nerve‐mediated ATP transient followed the tension transient (Figure 1a) which is counterintuitive if ATP release generates force. However, Figure  S2 shows that this delay of the ATP response can be explained by delays in the ATP‐electrode response.

The ability to attenuate selectively nerve‐mediated release of ATP rather than that of ACh offers the possibility of an interesting therapeutic target, as atropine‐resistant contractions occur only in tissue from human‐DO bladders. Adenosine, acting via A₁ receptors, and the https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1304 inhibitor https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=4743 both attenuate the consequences of nerve‐mediated ATP release (Chakrabarty et al., 2019; Pakzad et al., 2016; Searl et al., 2015). However, adenosine also reduces ACh release (Silva‐Ramos et al., 2015), and the role of sildenafil in this context requires evaluation. A₁ receptor agonists and PDE5 inhibitors respectively reduce https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2352 or increase https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2347 intracellular levels, both of which can attenuate activity of https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=533 or https://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=532 ²⁺ channels (Fukuda et al., 1996; Grassi, D'Ascenzo, & Azzena, 2004; Nickels et al., 2007) that mediate Ca²⁺ influx necessary for vesicular neurotransmitter release. Of interest is that https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2536, an N‐type blocker, but not https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2529, a P/Q‐type blocker, attenuates nerve‐mediated nucleotide release in human bladder (Breen, Smyth, Yamboliev, & Mutafova‐Yambolieva, 2006). However, further studies are required to identify more clearly which Ca²⁺ channel subtypes regulate vesicular ATP release and if they differ from ACh release pathways.

4.4. Limitations

The small size of the human biopsy samples precluded measurement of E‐NTPDase subtypes by western blot as well as allowing tension measurements with a separate strip. Thus, we relied on qPCR to provide transcription data. ARL‐67156 is a weak E‐NTPDase inhibitor, in particular, for the subtype most transcribed in detrusor, E‐NTPDase‐1 (Lévesque et al., 2007). However, it may underestimate ecto‐ATPase activity as a proportion of total tissue ATPase activity, it is assumed the proportional underestimation is similar in all preparation cohorts. In addition, it is assumed that ABMA desensitises P2X1/3 receptors and not any other subtypes that may generate detrusor contractile activity. It has been suggested that with experiments using variable stimulation frequencies to elicit nerve‐mediated contraction, it is not the frequency per se but the number of stimuli that determines contraction magnitude. This has been addressed in Figure S3, which shows that frequency of stimulation does indeed seem to be the relevant variable. Figure S3 is an account of why the different stimulus parameters have been chosen to generate contractions.

AUTHORS CONTRIBUTIONS

C.H.F. and C.Mc.C. devised the study. C.Mc.C., Y.I., D.S., B.C., and C.H.F. contributed to the experiments. C.H.F., C.Mc.C., A.J.K., and R.I.J. wrote drafts of the manuscript. All authors approved the final manuscript.

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for https://bpspubs.onlinelibrary.wiley.com/doi/full/10.1111/bph.14207 and https://bpspubs.onlinelibrary.wiley.com/doi/full/10.1111/bph.14206, and as recommended by funding agencies, publishers and other organisations engaged with supporting research.

Supporting information

Figure S1a Nerve‐mediated contractions in the absence and presence of 1 μM atropine. Data are shown for, A: human‐stable (n = 19), B: human‐DO (n = 16) and C: guinea‐pig (n = 7) preparations. Connecting lines for humanstable data are not shown for clarity. Median values in each data set are shown by the horizontal bars. p‐values are from paired Wilcoxon tests.

Figure S1b. A: atropine resistance in human‐stable (n = 19), human‐DO (n = 16) and guinea‐pig (n = 7) preparations. B: percentage reduction of nerve‐mediated contractions by apyrase in human‐stable, human‐DO and guinea‐pig preparations. Median values in each data set are shown by horizontal bars. p‐values are from Wilcoxon tests.

Figure S1c. Nerve‐mediated contractions (8 Hz) in the absence and presence of 1 μM a,b methylene‐ATP (ABMA). Data are shown for, A: human‐stable (n = 14), B: human‐DO (n = 11) and C: guinea‐pig (n = 11) preparations. The lines connect corresponding tension values in the absence then presence of ABMA in the same preparation. Mean values in each data set are shown by the horizontal bars. p‐values are from paired ttests. The inset by the guinea‐pig data shows a record of nerve‐mediated contractions in the presence of ABMA.

Figure S1d. Percentage reduction of nerve‐mediated contractions by apyrase in human‐stable, human‐DO and guinea‐pig preparations. Median values in each data set are shown by horizontal bars. p‐values are from Wilcoxon tests.

Figure S1e. A: ABMA pEC50 values from concentration‐response curves to generate contractions from detrusor strips (left) or intracellular Ca2+ transients from isolated myocytes (right) from human‐stable (n = 7), human‐DO (n = 7) and guinea‐pig (n = 7) detrusor. B: Efficacy of ABMA (1 μM) to generate a contraction from human‐stable (n = 15), human‐DO (n = 15) and guinea‐pig (n = 9), detrusor preparations. Mean values in each data set are shown by horizontal bars.

Figure S1 f. ATP pEC50 values from concentration‐response curves to generate contractions from detrusor strips from human‐stable (n = 20), human‐DO (n = 16) and guinea‐pig (n = 8) detrusor. Mean values in each data set are shown by horizontal bars. p‐values are from unpaired t‐tests.

Figure S1 g. ARL‐67156 sensitive values of Vmax (part A) and Km (part B) of tissue preparations from human‐stable (n = 8), human‐DO (n = 11) and guinea‐pig (n = 8) detrusor. Mean values in each data set are shown by horizontal bars. p‐values are from unpaired t‐tests.

Figure S1 h. Combined ENTPDase transcription for ENTPDase‐1, ENTPDase‐2, ENTPDase‐3 and ENTPDase‐5 from samples of human‐stable (n = 9) and human‐DO (n = 9) detrusor. Mean values of the ENTPDase‐1 data sets are shown by horizontal bars. p‐value are from an unpaired t‐test.

Figure S1i. f1/2 values of nerve‐mediated contractions (left, n = 18) and ATP‐transients (right, n = 6) before and after 1 μM atropine. Lines connect data from individual experiments. p‐values are from a paired ttest. The tension data points with the small white dots correspond to experiments from which ATP data were also collected.

Figure S2. Characterisation of ATP electrode responses. The ATP electrode response invariably lagged behind the tension recording which is counter‐intuitive if the tension transient was in response to the ATP transient. One explanation is a delay introduced by the ATP electrode itself. A: Electrodes were placed in the same tissue bath as used for experiments and calibrated with a rapid increase of ATP by microinjection, a typical response is shown. The rising phase was fitted by an increasing exponential function and with six electrodes a time constant, t, of 1.68 ± 0.66 s was measured. B: a model incorporating a tension transient that peaked after 2.4 s, a modelled real‐time increase of ATP for the same period, with an on/off t of 0.2 s. The ATP electrode response with an additional t of 1.68 s is slower, peaks at a lower response and shows a lag relative to the tension peak.

Figure S3a. Recordings of isometric tension recorded by stimulation with a 3‐­‐s pulse train, with 100 μs stimuli delivered at 8 Hz. Recordings were made in Tyrode's solution at 36°C in the absence (red trace) or presence (blue trace) of 1 μM tetrodotoxin.

Figure S3b. Recording of an isometric contraction (red trace) in isolated detrusor smooth muscle elicited by a 3‐­‐s stimulus train of 100 μs pulses delivered at 8 Hz. The black trace is a differential (dT/dt) of the contraction, achieved by analogue differentiator.

Figure S3c. Upper panels: isometric recordings of contractions generated with pulse trains of fixed duration and varying frequency (left) or varying duration and fixed frequency (right). In each case the number of stimuli in each train are shown in parentheses. Data are from the same preparation carried out as consecutive experiments. Lower panels: left, plots of tension vs number of stimuli for the two protocols of the above experiment. The two data points generated by a 3‐‐‐s, 8 Hz frequency for each protocol are arrowed. Lower panels; right, plots of seven similar paired experiments with equal numbers of stimuli in either protocol. Contractions are normalised to the average of data from the two protocols at using 24 stimuli (3‐‐‐s train at 8 Hz). Data are mean ± SE (n = 7). Curves are best‐‐‐fits of T = (T _max * s ^s)/(s ⁿ + $s_{\mathrm{1/2}}^n$), where Tmax is estimated maximum tension at high numbers of stimuli, s; $S_{\mathrm{1/2}}$. is the number of stimuli requited to generate Tmax/2; n is a constant.

Figure S3d. Left: the decline of contraction magnitude in 1 μM atropine (3‐‐‐s pulse train at 8 Hz). The curve is a fit to, T = (A.esp(− t/τ) + B) where A,B are constants, τ = time constant. Right: the percentage remainder of contraction amplitude as a function of stimulus number, as varied by altering frequency for a fixed duration stimulus (blue curve) or altering pulse‐‐‐train duration at a fixed frequency (black curve). Data, mean ± SE, n = 10 for variable frequency, n = 7 for variable duration.

Figure S3e. Force recorded in a detrusor preparation when stimulated at a constant number of stimuli (10 or 20) delivered at varying frequencies.

McCarthy CJ, Ikeda Y, Skennerton D, et al. Characterisation of nerve‐mediated ATP release from bladder detrusor muscle and its pathological implications. Br J Pharmacol. 2019;176:4720–4730. 10.1111/bph.14840