Differential expression and function of nicotinic acetylcholine receptors in the urinary bladder epithelium of the rat

Abstract

It has been previously determined that the epithelial lining of the urinary bladder, or urothelium, expresses two subtypes of nicotinic acetylcholine receptors (nAChRs) that mediate distinct physiological effects in vivo. These effects include inhibition of bladder reflexes through α7 receptors and an excitation of bladder reflexes through α3-containing (α3*) receptors. It is believed that urothelial receptors mediate their effects through modulating the release of neurotransmitters such as ATP that subsequently influence bladder afferent nerve excitability. Therefore, we examined the distribution of nAChRs in the urothelium, as well as their ability to influence the release of the neurotransmitter ATP. Immunofluorescent staining of both whole bladder tissue and primary urothelial cultures from the rat demonstrated that the urothelium contains both α3* and α7 receptors. In primary urothelial cultures, α7 stimulation with choline (10 μm to 1 mm) caused a decrease in basal ATP release while α3* stimulation with cytisine (1–100 μm) caused a concentration-dependent, biphasic response, with low concentrations (1–10 μm) inhibiting release and higher concentrations (50–100 μm) increasing release. These responses were mirrored in an in vitro, whole bladder preparation. In vivo, excitation of bladder reflexes in response to intravesical cytisine (100 μm) is blocked by systemic administration of the purinergic antagonist PPADS (1 or 3 μg kg⁻¹). We also examined how each receptor subtype influenced intracellular Ca²⁺ levels in cultured urothelial cells. nAChR stimulation increased µCa²⁺½_i through distinct mechanisms: α7 through a ryanodine-sensitive intracellular mechanism and α3* through extracellular influx. In addition, our findings suggest interactions between nAChR subtypes whereby activation of α7 receptors inhibited the response to a subsequent activation of α3* receptors, preventing the increase in µCa²⁺½_i previously observed. This inhibitory effect appears to be mediated through protein kinase A- or protein kinase C-mediated pathways.

Key points

• It has been previously shown that stimulation of urothelial nicotinic acetylcholine receptors (nAChRs) can alter reflex bladder activity in the rat; the current study examines this further.

• Stimulation of rat urothelial cells with an α7 nAChR agonist increases intracellular calcium through internal stores and decreases basal ATP release.

• Stimulation with an α3* nAChR agonist increases intracellular calcium through extracellular influx and increases basal ATP release.

• Infusion of an α3* agonist into the bladder lumen of the rat increases reflex bladder activity, which is blocked by intra-arterial administration of a purinergic antagonist.

• The cellular effects of α3* stimulation are blocked when the cells are pretreated with an α7 agonist, suggesting cross-talk between the receptors: this cross-talk may be mediated through protein kinase A or protein kinase C.

Introduction

Neuronal nicotinic acetylcholine receptors (nAChRs) have long been known to have many physiological roles throughout the body. For example, they are the primary receptors responsible for synaptic transmission in autonomic ganglia (De Biasi, 2002). nAChRs are also present in the neurons of the central nervous system, where they are being studied for their role in nociception (Damaj et al. 1998; Kesingland et al. 2000), cognition (Dajas-Bailador & Wonnacott, 2004; Singh et al. 2004), and memory (Chan et al. 2007). A new role for neuronal nAChRs, however, has begun to be elucidated: their role in the physiology of non-neuronal tissues (Conti-Fine et al. 2000). For example, nicotinic receptors of neuronal subtype (i.e. α3, α4 or α7 receptors) have been found in skin keratinocytes (Kurzen et al. 2007), bronchial epithelial cells (Wessler & Kirkpatrick, 2001), colonic epithelial cells (Summers et al. 2003), vascular endothelial cells (Macklin et al. 1998), and numerous immune cells such as macrophages, mast cells and mononuclear leukocytes (Gwilt et al. 2007). In these cell types, nAChRs mediate a wide variety of physiological functions, such as cell proliferation and differentiation, cell motility, modulation of endothelial permeability and release of chemokines/cytokines (Wessler et al. 1998).

We have previously reported that the epithelial lining of the urinary bladder (urothelium) also expresses nAChRs and that stimulation of these receptors through intravesical administration of agonists could significantly alter functional voiding activity of the rat (Beckel et al. 2006). Specifically, the urothelium expresses two types of nAChR: (1) α7 homomeric receptors that inhibit bladder reflexes, and (2) α3-containing (α3*) heteromeric receptors that have an excitatory effect on bladder reflexes. It has been hypothesized that these receptors mediate their effects by modulating the release of transmitters from the urothelium. These transmitters are thought to then act on afferent nerve terminals in the bladder wall to (directly or indirectly) influence bladder activity (Andersson, 2002). The plausibility of this hypothesis is supported by a number of different studies that demonstrate that the urothelium releases such neurotransmitters as ATP, ACh and nitric oxide (NO) in response to physical and chemical stimulation (for review see Apodaca et al. 2007; Birder, 2010). In terms of cholinergic receptors, the urothelium also expresses all five muscarinic acetylcholine receptors (M1 to M5) (Bschleipfer et al. 2007; Zarghooni et al. 2007; Kullmann et al. 2008a) and stimulation of these receptors in vitro results in the release of mediators including ATP and NO, which can impact bladder function (Beckel et al. 2004; Kullmann et al. 2008b). There is much less information on the involvement of nicotinic receptors and signalling events that may lead to changes in reflex voiding function.

The present study was undertaken to examine the involvement of urothelial nAChRs in the release of a transmitter, specifically ATP, and if the excitation of bladder reflexes observed in vivo following α3* nAChR stimulation could be blocked through antagonists against purinergic receptors. Because ATP release from the urothelium is thought to be a calcium-dependent process (Birder et al. 2003), we also examined each nAChR subtype's effect on intracellular calcium signalling, in an attempt to further elucidate the cellular mechanisms responsible for the effects of nAChR stimulation observed in vivo. Our results indicate that stimulation of different nicotinic receptors in the urothelium modulates ATP release in a reciprocal manner, with α7 stimulation inhibiting ATP release and α3* stimulation decreasing or increasing basal ATP release, depending on agonist concentration. The mechanism for these effects may involve calcium-dependent intracellular signalling pathways, as stimulation of the two subtypes of nicotinic receptor also increased intracellular calcium through distinct mechanisms. Finally, our research indicates a previously unreported interaction between nAChRs, as activation of α7 receptors can inhibit the µCa²⁺½_i transients normally elicited by α3* stimulation.

Methods

Animals

All experiments used female Sprague–Dawley rats (250–300 g) that were fed a standard diet with free access to water prior to experimentation. A total of 60 rats were used for the experiments described. Tissues were harvested from rats that were killed by inhalation of 100% CO₂ followed by cervical dislocation. All studies were carried out with the approval of the University of Pittsburgh Institutional Animal Care and Use Committee and maintained according to the standards set forth in the American Physiological Society's Guide for the care and use of laboratory animals.

Reagents and solutions

All nicotinic agonists and antagonists used in this study were obtained from Tocris Bioscience (Ellisville, MO, USA). All other reagents were obtained from Sigma-Aldrich Inc. (St Louis, MO, USA), excepting cell culture reagents, which were obtained from Invitrogen, Inc. (Carlsbad, CA, USA). The composition of Hanks’ balanced salt solution (HBSS) used during ATP and calcium experiments was (in mm): KCl, 5.0; KH₂PO₄, 0.3; NaCl, 138; NaHCO₃, 4.0; Na₂HPO₄, 0.3; glucose, 5.6; CaCl₂, 2.0; MgCl₂, 1.0; Hepes, 10.0. In experiments in the absence of extracellular calcium, CaCl₂ was omitted, the concentration of NaCl was increased to 140 mm and 0.5 mm EGTA was added. The composition of the Krebs solution used for the whole bladder preparation was (in mm): NaCl, 113; KCl, 4.7; MgSO₄, 1.2; NaHCO₃, 25.0; KH₂PO₄, 1.2; glucose, 11.5; CaCl₂, 2.5; Hepes, 10.0. All solutions were adjusted to pH 7.4 with NaOH and to 300 mosmol l^-1 using mannitol.

Rat urothelial cell culture

Rat urinary bladders were excised and immediately placed in Dulbecco's modified Eagle's medium (DMEM). The bladder was cut open, gently stretched with the epithelial side up and pinned in a Sylgard-coated dish. The bladder was incubated overnight at 4°C in DMEM, augmented with penicillin/streptomycin/fungizone and 2.5 mg ml⁻¹ dispase. The epithelium was then gently scraped from the underlying tissue using a spatula, placed in a culture flask and treated with enzyme (0.25% trypsin) to dissociate urothelial cells. Following dissociation, the cell suspension was placed in minimal essential medium containing 10% fetal bovine serum and centrifuged at 416 g for 15 min. The cells were resuspended in 1 ml of keratinocyte medium and 0.1 ml of the cell suspension (50,000–250,000 cells ml⁻¹) was added to the surface of collagen-coated glass coverslips for Ca²⁺ imaging or collagen-coated 35 mm plastic culture plates for ATP release studies. Cells were used within the first 1–2 days following plating.

Immunofluorescence staining

For whole bladder staining, bladders were removed, cut open longitudinally and snap frozen in OCT compound using liquid nitrogen. These bladders were then sectioned at 6 μm, placed on slides and air dried. For cultured cells, cells were grown on collagen-coated glass coverslips for 48 h prior to fixation. Slides or coverslips were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min. Following washes in PBS, the slides were incubated in 0.3 m glycine for 20 min to reduce auto-fluorescence. The slides were then permeabilized in 0.1% Triton X-100 and then blocked for 30 min. The blocking solution contained 0.1% Triton X-100, 0.5% bovine serum albumin and 10% normal donkey serum. The slides were then incubated with Alexa-488 α-bungarotoxin (for α7 receptors, 1 μm) or goat anti-α3 polyclonal antibody (1:100 dilution) for 2 h at 25°C. After primary incubation, α3 slides were washed with PBS and incubated with a donkey anti-goat-fluorescein isothiocyanate (FITC) secondary antibody (1:250 in PBS) for 1 h. For co-localization studies, slides were also incubated with mouse monoclonal antibodies against either cytokeratin 17 or 20 (1:50 dilution each) followed by a donkey anti-mouse-Cy3 (1:250) secondary antibody. All slides were also incubated with 4′,6-diamidino-2-phenylindole (DAPI; 1:5000 in PBS for 5 min) to provide a nuclear counterstain. Following final washes with PBS, the slides were mounted and viewed. Results shown are representative of experiments performed from bladders of three separate animals. Antibodies for the cytokeratins were obtained from Dako; the anti-α3 polyclonal antibody, Alexa-labelled α-bungarotoxin and all secondary antibodies were obtained through Invitrogen.

Calcium imaging (Fura-2)

Urothelial cells cultured on collagen-coated coverslips for 24–48 h were washed with HBSS then incubated with the acetoxymethyl ester form of Fura-2 (Fura-2 AM; 5 μm, 30 min incubation at 37°C). The cells were maintained in HBSS throughout the experiment by use of a gravity-fed perfusion system at a flow rate of approximately 2.4 ml min⁻¹. To record changes in the Fura-2 ratio, the cells were alternately illuminated at 340 and 380 nm using a xenon arc lamp and imaged at 510 nm with a Dage-MTI cooled CCD camera with 640 × 480 pixel resolution. A Dage-MTI Gen. II system image intensifier and software package (Compix Inc., Cranberry, PA, USA) was used to collect data. The cells were then stimulated by switching the perfusate to a nicotinic agonist and samples taken every 5 s for the duration of stimulation (typically 2 min). For antagonist studies, the appropriate antagonist was perfused for 10 min prior to as well as during agonist stimulation. In experiments examining the effects of kinases on nAChR-mediated calcium signals, protein kinase A (PKA)/protein kinase C (PKC) agonists and antagonists were perfused for 15 min prior to and during nicotinic receptor stimulation. Before the experiment, each cell in a field was denoted as a region of interest (ROI). Background fluorescence was also marked by a ROI that contained no cells. To calculate the Fura-2 ratio, the fluorescence intensity was averaged across each ROI and the background subtracted before the ratio of fluorescence (340/380) was taken. To calculate the peak change in the Fura-2 ratio following drug application, the baseline calcium signal, as determined by averaging the five readings immediately prior to drug application, was subtracted from the highest signal observed during the drug application. Increases in the Fura-2 ratio that were less than 3-fold greater than the standard deviation of the mean of the baseline samples were eliminated from the analysis, as they were within the range that could be attributed to noise during the recording.

ATP release

Urothelial cells were cultured as described above, plated on 35 mm plastic culture dishes and utilized within 24–48 h following plating. During the experiment, cells were perfused with HBSS using a peristaltic pump (flow rate: 0.6 ml min⁻¹) and perfusate (100 μl) was sampled at 30 s intervals. Samples were taken for 5 min prior to agonist perfusion, for 5 min during agonist perfusion and for 5 min during washout of the agonist with HBSS. For antagonist studies, samples were taken during the initial HBSS perfusion, for 5 min during antagonist perfusion, for 5 min during perfusion of antagonist and agonist combined and for 5 min during washout of the drugs using HBSS. ATP levels were quantified immediately following sampling using the luciferin–luciferase reagent (ATP Luminescence Kit, Sigma-Aldrich) and bioluminescence measured using a luminometer (Turner Designs, TD 20/20 Sunnyvale, CA, USA). Relative luminescence units were converted to ATP concentrations using a set of standards with known concentrations of ATP (dynamic range: 1 pm to 1 μm). To eliminate the possibility that changes in ATP measurements were due to interference of the drug with the luciferin–luciferase reaction the standards were also run in the presence of each agonist and antagonist at the highest concentration used in our experiments. Data are reported as a change in basal ATP release, measured as the difference in ATP concentration (in pm) between the average of five readings taken immediately prior to perfusion and the average of five readings taken during perfusion of a drug.

For whole bladder studies, bladders were removed, cut open longitudinally, affixed in a tissue chamber, and covered with Krebs solution. During the experiment, the tissue was perfused with Krebs solution using a gravity-fed system (flow rate: ∼3.0 ml min⁻¹) and perfusate (50 μl) was sampled from the mucosal surface at 30 s intervals. Each sample was added to a microcentrifuge tube containing 50 μl of 200 μmβ,γ-methylene ATP (an ATPase inhibitor) on ice. At the end of each experiment the samples were frozen at −80°C until reading. Samples were taken every 30 s for 5 min prior to agonist perfusion, for 5 min during agonist perfusion and for 5 min during washout of the agonist with Krebs solution. For antagonist studies, samples were taken during the initial Krebs perfusion, for 5 min during antagonist perfusion, for 5 min during perfusion of antagonist and agonist combined, and for 5 min during washout of the drugs using Krebs solution. ATP levels were quantified using the luciferin–luciferase reagent and analysed as previously described, correcting for the 2-fold dilution.

In vivo bladder cystometrogram

Rats (n= 4) were anaesthetized with urethane (1.2 g kg⁻¹, s.c.). After confirming a surgical plane of anaesthesia, a midline incision was made and a catheter (Intramedic tubing, PE 50) was inserted into the bladder lumen through a small incision in the bladder dome and secured with a purse-string suture. The catheter was connected by way of a three-way stopcock to a syringe pump for fluid infusion and a pressure transducer connected to a computer to record changes in bladder pressure. The ureters were also tied and cut to prevent filling of the bladder from the kidneys. A moistened wick of sterile gauze was inserted into the midline incision and sewn in place to drain urine from the abdominal cavity. Additionally, a catheter (Intramedic tubing, PE 10) was placed into the femoral artery to allow intra-arterial administration of drugs. Saline was infused intravesically (0.04 ml min⁻¹) for a control period (generally 1–2 h) to establish baseline bladder activity. Following the control period, the infusate was switched to a saline solution containing 100 μm cytisine and recordings taken for an additional hour. The purinergic antagonist pyridoxalphosphate-6-azophenyl-2′,4′-disulfonic acid (PPADS; 1 or 3 μg kg⁻¹) was then injected intra-arterially and bladder activity recorded for 10 min following drug administration.

Statistical analysis

Statistical significance was assessed in all experiments using either unpaired, two-tailed Student's t tests or one-way ANOVA with Tukey's multiple comparison post test. Statistical significance was accepted when P < 0.05.

Results

Urothelial cells express nAChRs, both in vivo and in vitro

In order to localize each receptor subunit to a specific location in the bladder, we performed co-localization studies in rat bladder tissue with cytokeratins 17 and 20, which are known to be differentially expressed in the urothelium of the rat. Cytokeratin 20 has been shown to be expressed in the umbrella cells of the urothelium (Schaafsma et al. 1989), while cytokeratin 17 localizes to the intermediate and basal cells (Troyanovsky et al. 1992).

In rat bladder tissue, the α3 subunit stained a layer of tissue consistent with the urothelium, as the staining was prevalent in tissue directly adjacent to the lumen of the bladder (Fig. 1). This staining disappeared if the antibody was pre-incubated with its antigen or if only the secondary antibody was used (data not shown). Co-localization studies with cytokeratin 20 suggested that the α3 receptor is expressed mainly in the umbrella cells (Fig. 1A–C), as the two antibodies co-localize almost exclusively.

Figure 1. Fluorescent staining of bladder tissue and cultured urothelial cells for nAChRs.

A–I, co-localization of nAChRs with the urothelial-specific markers cytokeratin 17 (intermediate and basal cells) and cytokeratin 20 (umbrella cells). The left column depicts nAChR staining, the middle column depicts cytokeratin and nuclear DAPI staining, and the right column depicts a merged image to show co-localization. The lumen of the bladder is denoted by the letter ‘L’. All micrographs were taken with a ×20 objective; the calibration bars represent 50 μm. A–C, α3 nAChR staining (green) co-localizes with cytokeratin 20 (red), indicating expression of the receptor in umbrella cells. Some staining was also seen throughout the bladder in smooth muscle (denoted by the * in C), and in suburothelial nerve fibres. D–F, α7 staining (green) co-localizes with cytokeratin 20 (red), indicating expression of the receptor in umbrella cells. G–I, α7 staining (green) co-localizes with cytokeratin 17 (red), indicating expression of the receptor in basal and intermediate cells. J–L, nAChR staining in primary cultures of urothelial cells, grown for 48 h. J, α7 staining. K, α3 staining. L, a merged image together with a DAPI nuclear stain (blue). Micrographs taken with a ×40 objective; the calibration bars represent 25 μm.

To localize the α7 nAChR subunit, a fluorescently tagged (AlexaFluor 488) epitope of the neurotoxin α-bungarotoxin (α-BTX) was used instead of an antibody. This was done because commercially available antibodies against α7 receptors have not yet been fully developed and α-BTX binds α7 receptors with high affinity and high specificity. α-BTX binding in rat bladder sections revealed staining consistent with urothelial localization (Fig. 1D and G), as staining was prevalent in tissue surrounding the bladder lumen. α-BTX staining in the rat bladder was diminished when pre-incubated with a 100-fold higher concentration of unlabelled α-BTX (data not shown). In order to determine which layers of the urothelium express α7 receptors, we co-stained rat bladder sections with α-BTX and cytokeratins 17 and 20. α-BTX staining co-localized with both cytokeratin 17 (Fig. 1D–F), suggesting α7 expression in intermediate and basal cells, and with cytokeratin 20 (Fig. 1G–I), suggesting α7 expression in umbrella cells.

In addition to examining nAChR expression in bladder tissue, we also examined expression in vitro using primary urothelial cultures. As shown in Fig. 1J–L, cultured cells expressed both α3 and α7 subunits. This staining was observed throughout all cells in the culture, and was located both cytoplasmically as well as along the cell membrane.

Stimulation of nAChRs modulates ATP release from primary cultures of rat urothelial cells and from whole rat bladders

A small (but quantifiable) release of ATP can be induced by fluid flow (HBSS at 0.6 ml min⁻¹) (average release: 20–40 pmFig. 2A). We first examined the effects of nAChR-subtype-specific agents had on this flow-induced release of basal ATP. Basal release of ATP could be inhibited in a concentration-dependent manner by the α7 agonist choline (Fig. 2A and B), decreasing the release of ATP on average by 55.4% at 1 mm choline. This choline-induced decrease in basal ATP release was reversible upon washout (8 min, Fig. 2A and B). To rule out any possibility of choline interfering with the luciferin–luciferase assay, we performed a standard curve with known concentrations of ATP, both in the presence of, and in the absence of, choline (Supplemental Fig. 1, available online only). These standard curves were identical, indicating that choline does not interfere with the assay. We used the α7 antagonist methyllycaconitine citrate (MLA; 100 μm, in HBSS) to determine if this decrease in ATP following choline was due to specific activation of α7 receptors. MLA blocked the effects of choline, and also increased basal ATP release when administered alone (Fig. 2C). MLA did not alter the standard curve, however, indicating that the increase in ATP release was not due to interference with the assay (Supplemental Fig. 1).

Figure 2. Choline inhibits ATP release from cultured urothelial cells.

A, representative trace of choline's effect on ATP release from cultured urothelial cells. B, summary of ATP experiments involving choline, expressed as a change in the concentration of ATP measured. These changes were calculated in each experiment as the difference between the average of 5 readings taken prior to and during choline perfusion. Each column summarizes the changes in 6 experiments, taken from 3 separate cultures. Normal HBSS perfused over cultured cells instead of choline (first column) in 6 experiments produced no significant change in baseline levels, suggesting that run down of urothelial ATP release under flow conditions was not a contributing factor. **P < 0.05 by one-way ANOVA followed by a Dunnett's post test to compare each column to the control (HBSS alone). C, effects of the α7 antagonist MLA on ATP release from urothelial cells. **P < 0.05 as compared by unpaired Student's t test.

Low concentrations of the α3* agonist cytisine (1–10 μm in HBSS, Fig. 3A and C) decreased ‘basal’ or flow-evoked ATP release. This decrease in the release of ATP (decreases of 14.7% for 1 μm cytisine, and 23.3% for 10 μm cytisine) is sustained for the length of drug application, but recovers almost immediately upon washout, unlike the prolonged decrease observed following choline (Fig. 3A). In contrast, higher concentrations of cytisine (50–100 μm, in HBSS) evoke a significant release of ATP above basal levels (Fig. 3B and C). This effect (an increase of 43.2% during perfusion of 100 μm cytisine) is also maintained for the duration of drug application and recovers with washout (Fig. 3B). To determine if the cytisine-induced effects were due to actions on α3* receptors and not through non-specific effects, we used the α3* antagonist 2,2,6,6-tetramethylpiperidin-4-yl heptanoate (TMPH). Perfusion of TMPH (90 μm, in HBSS) had no effect on ATP release alone, but did block the response to cytisine at both high and low concentrations (Fig. 3D). Cytisine did not interfere with the luciferin–luciferase assay, as we performed a standard curve with known amounts of ATP both with and without 100 μm cytisine. There was no difference in the standard curves, indicating that cytisine has no effect on the luciferin–luciferase assay (Supplemental Fig. 1).

Figure 3. Cytisine biphasically modulates ATP release from cultured urothelial cells.

A and B, representative trace of ATP release from urothelial cells in response to 10 μm (A) or 100 μm (B) cytisine. C, summary of cytisine's effects on ATP release from cultured urothelial cells. These changes were calculated in each experiment as the difference between the average of 5 readings taken prior to and during choline perfusion. Each column summarizes the changes in 6 experiments, taken from 3 separate cultures. **P < 0.05 by one-way ANOVA followed by a Dunnett's post test to compare each column to the control (HBSS alone). D, summary of the effects of the α3* antagonist TMPH on ATP release from cultured urothelial cells. **P < 0.05 as compared by unpaired Student's t test.

In addition to cultured cells, we examined if ATP release from intact urothelial tissue could be modulated by nicotinic agents. In a similar fashion to what was seen in cultured urothelial cells, choline (100 μm, n= 3) decreased basal release of ATP by 64.1% (Fig. 4), which was blocked by the α7 nAChR antagonist MLA (10.9% decrease from control, n= 3, P≤ 0.05). ATP release in response to cytisine also mirrored the results seen previously in cultured cells, with a lower concentration (10 μm) of the agonist decreasing basal ATP release (40.8%, n= 3, P≤ 0.05), while a larger concentration (100 μm) increased basal ATP release (33.2% over control, n= 3, P≤ 0.05). The increase in ATP release by cytisine was blocked by the α3* antagonist TMPH (100 μm, n= 3, P≤ 0.05).

Figure 4. Nicotinic agents modulate ATP release from whole bladder.

Summary graph depicting the effects of nicotinic agents on basal ATP release in the whole bladder preparation. These changes were calculated in each experiment as the difference between the average of 5 readings taken prior to and during drug perfusion. Each column summarizes the changes from 3–5 separate experiments each. **P < 0.05 as compared to control (Krebs solution only) by an unpaired Student's t test. ⁺⁺P < 0.05 between the two columns denoted by the accompanying line as compared by an unpaired Student's t test.

Excitation of bladder reflexes by intravesical cytisine can be blocked by PPADS

We used voiding cystometry to evaluate the effect of cytisine on reflex bladder activity in the anaesthetized rat. As shown in Fig. 5, the intercontraction interval (ICI) between voiding bladder contractions is decreased when cytisine (100 μm) is infused intravesically (0.04 ml min⁻¹), as compared to when saline was infused as a control using the same infusion rate. This facilitation was prevented when the purinergic antagonist PPADS (1 or 3 μg kg⁻¹, cumulative dose) is injected into the femoral artery, where the site of action is likely to be afferent nerve terminals located within the bladder wall. This inhibition of the bladder reflex is prolonged, lasting at least 15 min following injection.

Figure 5. Intra-arterial PPADS can block the excitation of bladder reflexes caused by intravesical cytisine.

A–C, representative traces of cystometry from saline-infused controls (A), 100 μm cytisine (B), and 100 μm cytisine after 3 μg kg^-1 PPADS (C). D, summary of intercontraction interval (ICI) time (seconds) during cystometrograms taken during saline infusion (control), intravesical administration of cytisine (100 μm), and following intra-arterial administration of PPADS (1 or 3 μg kg⁻¹ cumulative dose) concurrent with intravesical cytisine. n= 4 rats. *P < 0.05, **P < 0.005 as compared by unpaired Student's t test.

Activation of nAChRs increases intracellular calcium in cultured rat urothelial cells through distinct mechanisms

To examine the role of nAChRs in urothelial calcium signalling, we first stimulated α3* receptors in cultured rat urothelial cells with cytisine. As shown in Fig. 6, cytisine (1–100 μm, in HBSS) caused a concentration-dependent increase in the Fura-2 ratio (2.83 ± 0.54%, 9.35 ± 0.65% and 24.46 ± 0.80% increase over baseline for 1, 10 and 100 μm cytisine, respectively). These calcium transients slowly returned to baseline after approximately 30 s of continuous infusion, and could be completely blocked by a 10 min pre-incubation with the α3* receptor antagonist TPMH (30 μm, 3.74 ± 0.36% increase over baseline). Additionally, cytisine-induced calcium transients were blocked when stimulated in calcium-free HBSS containing 0.5 mm EGTA.

Figure 6. Stimulation of cultured urothelial cells with cytisine increases intracellular calcium.

A, representative traces of calcium imaging experiments following stimulation with 100 μm cytisine alone (continuous trace), cytisine following pre-incubation with α3* receptor antagonist TMPH (30 μm, dashed trace) and cytisine in HBSS containing no extracellular calcium (dotted trace). B, summary graph depicting the changes in intracellular calcium following cytisine stimulation as a change in the Fura-2 ratio (340/380). **P < 0.05 compared to HBSS control as determined by ANOVA with Tukey's multiple comparison post test. ‡P < 0.05 as compared to 100 μm cytisine by ANOVA with Tukey's multiple comparison post test. n= 15, 45, 37, 73, 72 and 115 cells, respectively.

To determine the contribution of α7 receptors to urothelial calcium signalling, cultured rat urothelial cells were stimulated with choline (1–100 μm, in HBSS). While these concentrations are in the lower range of the concentrations that can activate α7 receptors (EC₅₀, ∼400 μm), they were chosen in order to eliminate any effect of stimulation of muscarinic receptors, which may also be activated by higher concentrations of choline (concentrations greater than 100 μm). Additionally, these experiments were carried out in the presence of 10 μm atropine, a non-selective muscarinic receptor antagonist, to prevent muscarinic receptor activation. When choline was applied, a concentration-dependent increase in the Fura-2 ratio is observed (4.63 ± 0.19%, 14.25 ± 0.69% and 34.1 ± 3.30% increase over baseline for 1, 10 and 100 μm, respectively; Fig. 7A). This increase lasted for 3–5 min (Fig. 7A), and was significantly reduced by pre-incubation with the α7-specific antagonist α-bungarotoxin (1 μm in HBSS, 5 min incubation), indicating a specific action on the α7 receptor. Unlike cytisine, however, removal of extracellular calcium and the addition of EGTA to the bath did not block the choline-induced signal, suggesting the release of calcium from intracellular stores. Pre-incubation with ryanodine (10 μm in HBSS, 15 min) blocked this response. This concentration of ryanodine is generally accepted to block ryanodine receptor-mediated release of calcium from the endoplasmic reticulum (McPherson et al. 1991), thus indicating a role for ryanodine receptors in choline-induced release (Fig. 7B).

Figure 7. Activation of α7 nAChRs using choline increases intracellular calcium through release by intracellular stores.

A, representative traces of calcium imaging experiments following stimulation with 10 μm choline alone (continuous trace), choline following pre-incubation with α7 receptor antagonist α-BTX (1 μm, dashed trace) and choline in HBSS containing no extracellular calcium (dotted trace). B, summary graph depicting the changes in intracellular calcium following choline stimulation as a change in the Fura-2 ratio (340/380). **P < 0.05 compared to HBSS control as determined by ANOVA with Tukey's multiple comparison post test. ‡P < 0.05 as compared to 10 μm choline by ANOVA with Tukey's multiple comparison post test. n= 15, 30, 31, 40, 39, 35 and 92 cells, respectively.

α7 stimulation inhibits α3* receptor-mediated effects through PKC/PKA activation

To further examine the role of α7 nAChRs in eliciting changes in urothelial calcium, we used another selective agonist of α7 receptors, PNU 282987. PNU 282987 is a highly selective and potent α7 agonist (Bodnar et al. 2005), which would eliminate the need to include atropine in the bath to block potential non-selective activation of muscarinic receptors. Stimulation of cultured urothelial cells with PNU 282987 (10 nm to 1 μm, in HBSS) elicited no change in the Fura-2 ratio (Fig. 8A). However, it was noted that application of PNU 282987 (100 nm, 2 min pre-incubation) blocked any increase in the Fura-2 ratio observed following a subsequent stimulation with cytisine (100 μm, Fig. 8A). This inhibition of the α3*-mediated response by an α7 receptor agonist was blocked following pre-incubation with the α7 antagonist MLA (100 μm, Fig. 8B), but not if the antagonist was given after PNU 282987, indicating that the result was not due to actions downstream of α7 activation. It was also possible to recover the cytisine response if PNU 282987 was washed out of the bath for 10 min with normal HBSS prior to stimulation with cytisine.

Figure 8. Stimulation of α7 receptors blocks α3* receptor-mediated increases in intracellular calcium.

A, representative traces of increases in the Fura-2 ratio following either cytisine stimulation alone (100 μm, dashed line) or cytisine following PNU 282987 (100 nm, continuous line). Drug applications are denoted by the bars below the traces. Note that stimulation of cells with PNU 282987 alone elicited no change in the Fura-2 ratio. B, summary graph showing the change in Fura-2 ratio following stimulation of cultured urothelial cells with 100 μm cytisine alone and following stimulation with various α7 nAChR agents. The cytisine-induced signal is blocked following stimulation with the α7 agonist PNU 282987 (100 nm, 2nd column). This block is recoverable after 10 min wash with HBSS (3rd column) and is prevented if the cells are pre-incubated with the α7 antagonist methyllycaconitine citrate (MLA, 100 μm, 4th column). The inhibitory effect is not blocked, however, if the antagonist is given after the agonist (5th column). **P < 0.05 compared to 100 μm cytisine alone as determined by ANOVA with Tukey's multiple comparison post test. ‡P < 0.05 as compared to 100 μm cytisine after PNU by ANOVA with Tukey's multiple comparison post test. n= 73, 87, 87, 65 and 71 cells, respectively.

It has been shown that phosphorylation may result in inhibition of nAChRs (Nishizaki & Sumikawa, 1998a,b). Thus, we hypothesized that this inhibition of the α3*-mediated response may be due to phosphorylation of the α3* receptor. To test this hypothesis, we incubated urothelial cells with either PKA or PKC modulators and examined the influence on cytisine-induced calcium transients. Incubation for 15 min with either the PKC activator phorbol 12-myristate, 13-acetate (PMA, 100 nm in 0.1% DMSO) or Pseudo RACK1 (20 μm, in HBSS) blocked cytisine-induced Ca²⁺ signals (79.1% and 83.2% decrease from cytisine alone, respectively; Fig. 9A). DMSO, 0.1%, was also tested to rule out any vehicle effect on the Fura-2 ratio; no change in signal was observed (data not shown). Similar results were observed when using activators of PKA; 8-bromo-cAMP (30 μm, in HBSS) or dibutyryl-cAMP (1 mm, in HBSS) blocked increases in the Fura-2 ratio following cytisine stimulation (86.4% and 78.0% decrease from cytisine alone, respectively; Fig. 8B).

Figure 9. Cytisine-induced signals are modulated by PKC/PKA.

A, summary graph depicting the changes in intracellular calcium following cytisine stimulation in the presence of PKC activators (phorbol 12-myristate 13-acetate (PMA) and Pseudo RACK1) and inhibitors (chelerythrine chloride and Ro 32-0432) as a change in the Fura-2 ratio (340/380). **P < 0.05 compared to cytisine alone as determined by ANOVA with Tukey's multiple comparison post test. n= 73, 70, 84, 90 and 56 cells, respectively. B, summary graph depicting the changes in intracellular calcium following cytisine stimulation in the presence of PKA activators (8-Br-cAMP and dibutyryl-cAMP) and inhibitors (PKI 14-22 and PKA inhibitor (6-22)) as a change in the Fura-2 ratio (340/380). **P < 0.05 compared to cytisine alone as determined by ANOVA with Tukey's multiple comparison post test. n= 73, 62, 85, 65 and 74 cells, respectively.

We next examined whether kinase inhibition would increase cytisine-induced changes in intracellular Ca²⁺. Inhibition of PKC using either chelerythrine chloride (1 μm in HBSS) or Ro 32-0432 (1 μm, in 0.1% DMSO) potentiated the Ca²⁺ signal observed following cytisine stimulation (Fig. 9A). This potentiation was evident as an increase in the peak of the calcium response (20.0% and 193% increase over cytisine alone, respectively). Similarly, PKA inhibitors PKI 14-22 (100 nm, in HBSS) and PKA Inhibitor 6-22 (10 μm, in HBSS) also potentiated the cytisine-induced calcium signals (54.3% and 64.0%, respectively; Fig. 9B).

We also examined whether α7 and α3* interactions may involve protein kinase activation. Pre-incubation with either 1 μm chelerythrine chloride or 100 nm PKI 14-22 for 15 min reversed the α7 agonist PNU 282987-induced decrease in α3* (cytisine)-induced Ca²⁺ increase (Fig. 10). These data indicate that the inhibitory effects of PNU 282987 on α3*-mediated Ca²⁺ transients may involve activation of PKA or PKC.

Figure 10. Summary of the effects of the PKA inhibitor PKI 14-22 or the PKC inhibitor chelerythrine chloride on the PNU 282987-mediated inhibition of the calcium signals evoked by cytisine.

PNU 282987, 100 nm, blocks increases in the Fura-2 ratio following 100 μm cytisine (1st column), but pre-incubation of the cells with either PKI 14-22 or chelerythrine chloride removes that inhibition. **P < 0.05 as compared to 100 μm cytisine after PNU by ANOVA with Tukey's multiple comparison post test. n= 73, 87, 71 and 90 cells, respectively.

Discussion

Accumulating evidence supports the view that the urothelium is involved in sensory functions controlling micturition (Apodaca et al. 2007; Birder, 2010). Urothelial signalling mechanisms influencing bladder activity are likely to be complex, with release of the transmitters acetylcholine and ATP likely to impact smooth muscle contractility (Santoso et al. 2010) as well as neural excitability (Andersson, 2002). Both transmitters can be released by the urothelium in response to stretch or chemical stimulation (Yoshida et al. 2004, 2006; Hanna-Mitchell et al. 2007; Lips et al. 2007) though non-cholinergic mechanisms may play a more important role in pathology. In addition, though the urothelium expresses the full complement of receptors that are activated by ACh (i.e. muscarinic and nicotinic receptors) (Beckel et al. 2006; Bschleipfer et al. 2007; Zarghooni et al. 2007; Kullmann et al. 2008a), less is known regarding the role of urothelial nAChRs and involvement in bladder function.

In terms of urothelial muscarinic AChR stimulation, a recent study (Kullmann et al. 2008b) demonstrated that stimulation of urothelial muscarinic receptors using oxotremorine-M (OxoM) altered bladder reflexes in a manner dependent on ATP and NO release. Intravesical administration of OxoM (low concentrations) had an inhibitory effect on voiding reflexes which was prevented by the nitric oxide inhibitor l-NAME. In contrast, intravesical administration of higher concentrations of OxoM facilitated bladder activity, which was mediated in part by ATP as this response could be prevented by the purinergic antagonist PPADS. Our study also demonstrates a role for ATP in the modulation of bladder reflexes following activation of urothelial nicotinic receptors. Further, the facilitatory effect of cytisine on bladder activity is likely to involve ATP as PPADS could block this response. In addition, activation of α7 nAChRs inhibited the release of ATP from urothelial cells, and this decrease may play a part in the inhibitory effects of urothelial α7 nAChR stimulation in vivo which we previously reported (Beckel et al. 2006). We did not examine the role of additional transmitters such as NO, therefore it is also possible that NO production may also play a role in the inhibition of bladder reflexes observed following α7 nAChR stimulation.

OxoM elicits both a concentration-dependent inhibitory and excitatory effect on bladder reflexes in the anaesthetized rat. Though the mechanism has not been established, one possible explanation for this type of biphasic response may be due to general non-selectivity to muscarinic receptor subtypes present in urothelium. Hence, the inhibitory effects of OxoM could be due to activation of high-affinity receptors and the excitatory effects due to activation of low-affinity receptors. In the present study we demonstrated that cytisine can also elicit both an inhibitory as well as excitatory effect on ATP release from urothelial cells. It is also interesting to note the similar bi-phasic response the urothelium exhibits following stimulation with the α3* agonist cytisine. While α3-containing receptors are generally grouped together due to a lack of specific agonists or antagonists that could distinguish between them, there is evidence that the urothelium may express a number of α3* receptors, such as α3β4, α3α5β4, α3β3β4 and α3α5β3β4 (Beckel et al. 2006). When studied in heterologous expression systems such as oocytes, it is apparent that each receptor subtype has distinct pharmacological and electrophysiological properties that may elicit different physiological effects in the cell (Lindstrom et al. 1996; Gerzanich et al. 1998; Nelson & Lindstrom, 1999; Genzen et al. 2001). Therefore, we cannot rule out the possibility that the results we observe are not due to the activation of two (or more) separate subtypes of α3* receptor which have different affinities for the agonist cytisine: a high-affinity receptor that activates an inhibitory pathway and a low-affinity receptor that has an excitatory effect.

Our understanding of the mechanisms underlying the release of transmitters from urothelial cells is still relatively unexplored. In contrast to a role in non-neuronal cells, nicotinic receptor modulation of transmitter release from nerves has been fairly well characterized (for reviews see Wonnacott, 1997; Vizi & Lendvai, 1999; Marchi & Grilli, 2010). For example, nAChRs located presynaptically can modulate the release of a number of transmitters including noradrenaline (norepinephrine), glutamate, GABA and dopamine in the brain, spinal cord and periphery. One example in particular is that of presynaptic nAChR modulation of excitatory amino acid release from neurons cultured from the rat prefrontal cortex (Dickinson et al. 2008). These neurons also express two distinct classes of nAChR: α7 receptors and β2-containing receptors (presumably α4β2), which modulate the release of µ³H½aspartate (as a stand-in for glutamate) via distinct mechanisms. Release of µ³H½aspartate following activation of the β2* receptors was highly sensitive to removal of extracellular Ca²⁺ and to replacement of Ca²⁺ with the non-permeable ion Cd²⁺, suggesting a role for extracellular calcium and possibly coupling to voltage-operated calcium channels. However, release of µ³H½aspartate following the activation of α7 receptors was not blocked by Cd²⁺, and was blocked by ryanodine and xestospongin-C, blockers of intracellular calcium stores. Our data suggest that urothelial nAChRs also modulate intercellular calcium via two distinct methods: α3* receptors through extracellular influx and α7 receptors through release from intracellular stores. These data suggest some ‘neuron’-like properties of the urothelium, with expression of nAChRs and the ability to modulate (directly or indirectly) the sensory limb of the micturition reflex.

Given that an injection of PPADS into the femoral artery of the rat blocked the excitatory effects on ICI by intravesical cytisine, we have hypothesized that PPADS blocks the activation of P2X receptors present on bladder afferent nerve terminals. Our rationale behind this hypothesis comes from the change in the intercontraction interval; the ICI is generally believed to be influenced by changes in afferent nerve excitability. Therefore, since the ICI is decreased following systemic injection of PPADS, we assume that its site of action is on afferent nerves. This hypothesis is supported by earlier work by Yu & de Groat (2008), which examined the role of purinergic receptors in afferent nerve excitability using an isolated bladder–nerve preparation. In their experiments, PPADS applied to the bath maintaining the bladder decreased afferent nerve firing in response to stretch. This would be consistent with the results we have seen in vivo. However, purinergic receptors are also expressed in a number of locations throughout the bladder pathway, including the dorsal root ganglia (Ruan et al. 2005), bladder smooth muscle (Birder et al. 2004), and the spinal cord (Burnstock, 2001). Because we did not directly examine afferent nerve activity in our experiments, we must allow for the possibility that our data are the result of actions of PPADS at another location.

Possibly the most interesting finding of the present study is that stimulation of α7 receptors can turn off the excitatory effects of α3* receptors; α7 agonists can block the rise in intracellular Ca²⁺ and increase in ATP release exhibited by cytisine. These data reveal an as-yet unexplored mechanism of nAChR signalling: the modulation of one type of nicotinic receptor by another. Our research indicates that this mechanism may be mediated through a PKA/PKC-dependent pathway, presumably through phosphorylation of the α3* receptor leading to inactivation. These data are supported by previous studies linking nicotinic receptors to protein kinases. For example, it has been shown that stimulation of α7 nicotinic receptors can lead to the activation of PKC (Chernyavsky et al. 2004; Sun et al. 2004). Other studies have indicated that specific residues located in the M3/M4 cytoplasmic domain of α4β2 nicotinic receptors can be phosphorylated by either PKC or PKA (Pollock et al. 2007). This could lead to changes in receptor maturation and insertion into the plasma membrane (Jeanclos et al. 2001; Nashmi et al. 2003; Exley et al. 2006) as well as the receptor's electrophysiological properties, such as rate of desensitization and time to recovery from desensitization (Marszalec et al. 2005). Further experimentation must be performed, however, to determine if these sites exist in α3* receptors and if phosphorylation of these residues leads to the inhibition of the receptor that we demonstrate in the present study.

Our finding that urothelial nAChRs can play a role in the modulation of ATP release may indicate a role in nociceptive signalling in the lower urinary tract. For example, nAChR agonists have already been demonstrated to have analgesic properties, which are thought to be due to actions on nAChRs on the central terminals of afferent nerves, preventing the release of neurotransmitters (Ribeiro-da-Silva & Cuello, 1990; Khan et al. 2004) that activate secondary spinal nerves. Our research demonstrates that an α7 nAChR agonist can also block the release of ATP from the urothelium, which may be useful in the treatment of visceral disorders, such as painful bladder syndrome/interstitial cystitis (PBS/IC), where increased urothelial ATP release may play an important role in the symptomatology of this disorder (Cook & McCleskey, 2000; Birder et al. 2003). In addition, activation of α7 receptors results in inhibitory effects on voiding frequency (Beckel et al. 2006), suggesting an additional role for urothelial nicotinic receptors as a therapeutic target for overactive bladder disorders. This may be of additional clinical importance given the possible involvement of urothelial muscarinic receptors in the changes in bladder sensation and issues including non-selectivity and side effect profiles.

Glossary

Abbreviations

α-BTX

α-bungarotoxin

HBSS

Hanks' buffered salt solution

ICI

intercontraction interval

MLA

methyllycaconitine citrate

nAChR

nicotinic acetylcholine receptor

PKA

protein kinase A

PKC

protein kinase C

PPADS

pyridoxalphosphate-6-azophenyl-2′,4′-disulfonic acid

TMPH

2,2,6,6-tetramethylpiperidin-4-yl heptanoate

Author contributions

The conception and design of the experiments were done by both J.M.B. and L.A.B. The experiments were carried out in the laboratory of L.A.B., except for some minor experiments requested by reviewers, which were carried out in J.M.B.'s current laboratory at the University of Pennsylvania. Collection, analysis and interpretation of the data were performed by J.M.B. The manuscript was written by J.M.B. with editorial revisions by L.A.B.

Supplementary material

Supplementary Figure 1