The mechanism whereby the urinary bladder senses fullness is unclear. ATP is released from the serosal surface when the bladder is stretched (Ferguson et al. 1997). Furthermore, P2X₃ receptor-deficient knock-out mice exhibit urinary retention and their absence from sub-urothelial sensory afferents may underlie reduced sensation of bladder fullness (Cockayne et al. 2000). Recently a sub-urothelial myofibroblast layer has been described with cells that make close appositions to nerves (Wiseman et al. 2003) and are connected to each other through gap junctions (Sui et al. 2002). We have studied the electrophysiological properties of these cells and their response to ATP to determine if they could play a role in the sensory process.

Human bladder wall samples were obtained with Ethical Committee approval from patients undergoing either cystectomy or bladder augmentation. The urothelium was dissected from the underlying detrusor and digested at 37 °C with a collagenase-based medium. The tissue was partially disrupted into round urothelial cells and ovoid or spindle-shaped cells with or without dendrite-like structures. Experiments were performed at 37 °C in a HCO₃^–/CO₂-buffered superfusate. Electrophysiological recordings were made with patch-type electrodes filled with a KCl/EGTA-based solution. Intracellular [Ca²⁺] ([Ca²⁺]_i) was measured with the fluorochrome Fura-2. Data are presented as means ± S.D.

Cell capacitance was 27 ± 16 pF (n = 53) and membrane potential was -63 ± 13 mV (n = 16). However the membrane potential showed either spikes or small fluctuations. Membrane resistance was 8.5-9.0 X 10⁴ V cm² from resting potential changes to small depolarising or hyperpolarising currents. Under voltage clamp, on depolarisation from -100 mV sustained outward currents were preceded in some cells by a small, transient inward current. Inward current was attenuated in Ca²⁺-free superfusate and peaked at about -10 mV (25 ± 10 pA). Outward current showed outward-going rectification with spontaneous transient components, exhibited a reversal potential at about -80 mV and was greatly attenuated by 30 mM tetraethylammonium chloride. ATP at 30 µM increased [Ca²⁺]_i from 90 ± 60 nM to a peak of 832 ± 500 nM (18 transients, 4 cells). ATP was also applied to several epithelial cells, but no Ca²⁺ transients were observed. ATP at 100 µM generated, after a delay, transient inward currents when the membrane was clamped at -60 mV. The maximum peak inward current was 23 ± 17 pA (n = 6).

It is possible to isolate cells from the urothelial layer of the bladder, distinct from epithelial cells: they are not contaminant smooth muscle cells from the underlying detrusor but represent a distinct urothelial/suburothelial cell population, previously identified as myofibroblasts. The cells had many characteristics of excitable cells and responded to ATP so that they could form an electrical network to distribute an electrical signal over a reasonable area upon focal depolarisation. These properties are consistent with the hypothesis that these cells form an intermediate stage in the modality of bladder sensation and could act as a variable gain integrating stage.

We thank the Wellcome Trust for financial assistance