Excitatory neurotransmitters ATP and acetylcholine (ACh) are co-released from post-ganglionic nerve fibres to detrusor muscle in the mouse urinary bladder. Nerve-mediated contractions, i.e. labile to tetrodotoxin, are abolished by exposure to the metabolically stable ATP analogue, alpha,beta-methylene ATP (ABMA) which desensitises purinergic P2X receptors, and the non-selective muscarinic antagonist, atropine. The frequency-dependence of purinergic and cholinergic components of the contraction are different, with the purinergic component predominant at low stimulation frequencies in comparison to the cholinergic component at high stimulation frequencies [1]. This raises the possibility that ATP and ACh are differentially released from the nerve terminal and may be separately regulated. This is of interest because the purinergic component of contractions is only prominent in benign pathologies in the human bladder, such as overactive bladder associated with neuropathic, obstructive or idiopathic causes, and detrusor contractions from normal bladders are essentially completely cholinergic [2]. Current therapy includes the use of antimuscarinic agents, that suppress the physiological cholinergic component of contraction, and it would be attractive to develop agents that selectively suppress the purinergic component implicated in pathophysiology. The effect on neuronal transmitter release is inferred when using nerve-mediated contractions as the outcome measure in most studies, and more direct evidence would be desirable. Previous work has shown that differential suppression of low-frequency contractions is achieved by adenosine, acting through adenosine A1 receptor to inhibit cyclic AMP [3], or PDE5 inhibitors like sildenafil to elevate levels of cycle GMP [1]. The aim of this study was to test ATP and ACh concentrations during nerve-mediated stimulation of isolated detrusor preparations, with the hypothesis that measurement of nerve-mediated ATP and ACh release would provide a more direct test of the ability to differentially modulate release of co-transmitters.

Bladders were dissected from 12-week male C57BL/6 mice, and tissue strips were mounted to record contractions generated by electrical field stimulation (EFS; 0.1-ms pulses, 1-40 Hz, 3-s train every 90 s) and inhibited by tetrodotoxin. The tissue preparation was superfused with Tyrode’s solution (24 mM NaHCO3:5%CO2, pH 7.4, 36℃). Nerve-mediated ATP and ACh release was measured in superfusate samples (100 μl and 50 μl respectively) taken immediately before and during EFS. ATP was measured using a luciferin-luciferase assay, and ACh release was measured using a choline/ACh assay kit. Neurotransmitter release was plotted as a function of stimulation frequency and analysed to estimate the maximum release at high frequencies, and the frequency to generate half-maximal response, f1/2 (Hz). Release measurements were normalised for sample volume and wet tissue preparation weight. Drug interventions were delivered by the superfusate. Data are presented as mean±SD; differences between data sets were tested with repeated measures two-way ANOVA followed by a parametric post-hoc tests or Student’s paired t-tests where appropriate; the null hypothesis was rejected at p<0.05.

ACh release was measured over a range of stimulation frequencies between 1 and 40 Hz. The magnitude or release was described by a single function with a maximal value of 146±39 fmol.μl-1.mg wet weight-1 and a half-maximal frequency, f1/2, of 11.4±2.7 Hz (n=21, N=11; Figure 1A). The frequency-dependent range of release was limited, with 10-90% levels between 6 and 20 Hz (Figure 1B). No significant release above the background was recorded at 1 or 2 Hz stimulation. Release at all frequencies was completely abolished by lidocaine (2% final concentration) to the superfusate. 

ATP was measured initially over the same frequency range, but data were poorly-fitted to a single function. At the lowest frequency initially used (1 Hz), the percentage of ATP release compared to that estimated at high frequencies maximal output was 30.6±5.6% (n=93, N=51). A considerably better fit was obtained with a linear two-function fit suggesting that data could be modelled by two independent components with significantly different frequency-dependencies (Figure 2A). Both components had a similar capacity: the low- and high-frequency fractions were respectively 59±16 and 64±18 fmol.μl-1.mg wet weight-1 (p>0.05,n=93, N=51; Figure 2B). The frequencies for half-maximal release were significantly different: 0.86±0.16 and 15.1±4.5 Hz (p<0.0001, n=93, N=51). The total capacity of the two ATP pools (122±28 fmol.μl-1.mg wet weight-1) can be compared to that of the ACh pool (146±12 fmol.μl-1.mg wet weight-1). Overall 50% release of the total ATP pool was achieved at a significantly smaller stimulation frequency than that of the ACh pool: respectively 5.2±2.3 vs 11.4±2.7 Hz (n=93,21; N=51,11, p=5.10-7: Figure 2C). 

ACh and ATP release were also measured in the presence of 1 mM adenosine and 20 μM sildenafil to test if there was a differential effect on ACh and ATP release. ACh and ATP release is quoted at 20 Hz stimulation where output was maximal, and ATP release at 1 Hz is also quoted representing the low-frequency component. Neither adenosine or sildenafil had a significant effect on ACh release. Values in control and with adenosine were 1194±232 and 1235±232 fmol.μl-1 respectively (p=0.496, n=5, N=5). Respective values with control and sildenafil were 1004±111 and 975±140 fmol.μl-1 (p=0.132; n=5; N=5). By contrast, both adenosine and sildenafil reduced ATP release. Values at 1 Hz in control and adenosine were 34±6 and 21±2 fmol.μl-1 respectively (p<0.001; n=6; N=6), and at 20 Hz were 85±24 and 50±13 fmol.μl-1 respectively (p<0.001; n=6; N=6). For sildenafil, corresponding values at 1 Hz were 39±8 and 22±4 fmol.μl-1 respectively (p<0.001; n=8; N=8), and at 20 Hz were 104±22 and 59±14 fmol.μl-1 respectively (p<0.001; n=8; N=8).

These data suggest differential regulation for the release of the two neurotransmitters, revealing the potential for therapeutic agents to separately control their release. The curve for ACh release is best-fit to a one-component pool, whereas the curve for ATP release is best-fit to a two-component pool. Adenosine and sildenafil  have previously been shown to have an effect at low frequency nerve-mediated contractions [1, 3]. In this study, they had no significant effect on nerve-mediated ACh release. However both agents significantly reduced ATP release, with similar actions on the low- and high-frequency ATP components.

This study demonstrates the ability to measure neurotransmitter release and potential differential effects on ACh and ATP release. The differential reduction of ATP release by adenosine and sildenafil, but not ACh release, has important implications for their potential use in selectively attenuating contractions associated with human overactive bladder contractions. Current therapies target the normal physiological cholinergic pathway with the consequence that their effectiveness is limited. This study demonstrates the potential to specifically target the pathophysiological purinergic pathway implicated in bladder pathologies.

Chakrabarty B, Ito H, Ximenes M, Nishikawa N, Vahabi B, Kanai AJ, Pickering AE, Drake MJ, Fry CH (2019). Influence of sildenafil on the purinergic components of nerve-mediated and urothelial ATP release from the bladder of normal and spinal cord injured mice. Br J Pharmacol 176:2227-2237.
McCarthy CJ, Ikeda Y, Skennerton D, Chakrabarty B, Kanai AJ, Jabr RI, Fry CH (2019). Characterisation of nerve-mediated ATP release from bladder detrusor muscle and its pathological implications. Br J Pharmacol 176: 4720-4730.
Pakzad M, Ikeda Y, McCarthy C, Kitney DG, Jabr RI, Fry CH (2016). Contractile effects and receptor analysis of adenosine-receptors in human detrusor muscle from stable and neuropathic bladders. Naunyn Schmiedebergs Arch Pharmacol 389: 921-929.