To identify the cells expressing the M3 muscarinic receptor subtype in the lamina propria of the bladder.
To identify the cells expressing the M3 muscarinic receptor subtype in the lamina propria of the bladder.
The bladders from five normal guinea pigs were isolated and fixed in 4% paraformaldehyde. Tissues sections (10 µm) were then cut and stained with antibodies to the type 3 muscarinic receptor (M3), the interstitial cell marker vimentin and the nonspecific nerve marker PGP 9.5. The specificity of the antibody to the M3 receptor was established using the complementary blocking peptide and Western blot analysis of human embryonic kidney (HEK) cells transfected to express the M3 receptor protein.
The M3 antibody pre-incubated with its blocking peptide showed no immunohistochemical staining. Investigating this antibody using HEK cells transfected to express the M3 receptor protein and control HEK cells showed a single band in the transfected cells and no band in the control cells. M3 receptor immunoreactivity (M3-IR) was detected primarily on a dense network of vimentin-positive (vim+) cells lying immediately below the urothelium, i.e. the suburothelial interstitial cells (Su-ICs). The M3-IR was punctate and appeared to be located on the cell surface. The diffuse network of cells in the remaining regions of the lamina propria showed no M3-IR. Few nerve fibres were associated with the M3-IR Su-ICs. The M3-IR Su-ICs were most numerous and prominent in the lateral wall. The number of M3-IR/vim+ cells diminished towards the bladder base and were absent in the bladder urethral junction. In the base and urethral junction there were vim+ cells that were not M3-IR. A population of umbrella cells in the lateral wall also showed weak punctate M3-IR.
Using a well-characterized M3 antibody, these results show for the first time that the M3 muscarinic receptor in the lamina propria is located specifically on the Su-ICs. The physiological role of these cells is unknown and consequently the significance of what appears to be a cholinergic signalling system is unclear. Previously published data showed that these cells respond to nitric oxide and atrial natriuretic peptide with an increase in cGMP and possibly prostaglandin. All of these observations, taken together, suggest that the Su-ICs receive multiple inputs and that they must be part of a complex signalling system in this region of the bladder wall.
human embryonic kidney
Tris-buffered saline (Triton)
glyceraldehyde phosphate dehydrogenase
type 2 prostaglandin receptor
The importance of cholinergic muscarinic receptors during bladder stimulation was recognized over 100 years ago . It has long been known that the M3 receptors are responsible for activating the detrusor . This was shown elegantly using M3‘knockout’ mice where cholinergic activation of the bladder is drastically affected although the mice still appear to void normally [3,4].
One of the major clinical conditions associated with bladder dysfunction is a sensation of urgency during filling and an increased frequency of voiding. It was originally thought that these symptoms were associated with bladder contractions during the filling phase. Given the importance of M3 receptors in the activation of bladder contractions, M3 specific anticholinergic drugs were developed to treat the urological symptoms. Clinically, these drugs are effective in reducing symptoms of urgency and frequency. However, it has become apparent that, at therapeutic doses, these drugs do not affect either the nonvoiding contractions or the voiding contractions [5–7]. Thus, these drugs must also act on other systems in the bladder. This has led to the suggestion that the anticholinergic drugs act upon sensory mechanisms operating during the filling phase. The search is now on to find these mechanisms. Current thinking has suggested roles for cholinergic mechanisms associated with the urothelium, afferent nerves and the motor/sensory system [8–10].
Molecular studies have shown that the all five muscarinic receptors are expressed in the mouse  and human bladder [12,13]. However, it is generally accepted that only M2 and M3 receptors are involved in bladder smooth muscle contraction . The relative amounts of M2 or M3 receptors in the muscle and urothelial layers of several species, e.g. rat, mouse and human, are the not the same [11,12,15–18], the M2 receptor being present in greater amounts than the M3 receptor [16,19]. To unravel the different cholinergic elements in the normal and pathological bladder it is essential to know where the muscarinic receptors are located. The most obvious approach would be to use immunohistochemistry. However, the absence of specific and reliable antibodies has limited this route. Recent studies on the location of M2 and M3 receptors on the human bladder showed their presence either only on smooth muscle cells  or on both interstitial cells (ICs) in the lamina propria (LP) and muscle . Another study showed weak immunostaining of the human urothelium for the M2 receptor and strong M3 immunostaining in this tissue, whereas real-time PCR showed no difference in expression . Unfortunately, in these studies the specificities of the antibodies used was not thoroughly investigated. The aim of the present study was to investigate the location of the M3 receptor in the normal guinea pig bladder and in the bladder of animals with a previous created obstruction. A main concern of the study was to characterize several M3 antibodies, all marketed as specific reagents for this receptor subtype. Although it has been argued that the specificity of antibodies can never be confirmed [22,23], we used several tests on these antibodies. We transfected human embryonic kidney (HEK) cells with a rodent M3 receptor as a model system. In addition, we studied all antibodies by Western blotting and, where possible, with pre-absorption experiments. The antibody that passed all these tests was applied to sections of the guinea pig urinary bladder to localize the M3 receptor in this tissue.
Guinea pigs (five male, weight 270–300 g) were killed by cervical dislocation. All procedures were carried out in accordance with guidelines of the University of Maastricht and were in line with the European Community guidelines.
The urinary bladder, including the proximal urethra, was removed from each guinea pig and placed in ice-cold Krebs’ solution (mm: NaCl 121.1; KCl 1.87; CaCl2 1.2; MgSO4 1.15; NaHCO3 25; KH2PO4 1.17; glucose 11.0) and bubbled with 5% CO2 and 95% O2 (pH 7.4). Each bladder was divided into a ventral and a dorsal piece, and maintained in Krebs’ solution containing 1 mm of the nonspecific phosphodiesterase inhibitor isobutyl-methyl-xanthine (Sigma-Aldrich, Poole, Dorset, UK) at 36 °C for 30 min. Incubations were terminated by immersing bladder pieces in ice-cold fixative solution of 4% freshly prepared depolymerized paraformaldehyde for 120 min at 4 °C. Tissues were then fixed overnight at 4 °C in 0.1 m phosphate buffer with 10% sucrose and the next day the tissues were placed in 0.1 m phosphate buffer with 20% sucrose at 4 °C, and then washed overnight at 4 °C in 0.1 m phosphate buffer with 30% sucrose. The tissues were placed in OCT compound (Tissue-Tek, Miles, Elkhart, IN, USA) to form a single block. This was then snap-frozen in isopentane cooled in liquid nitrogen. Cryostat sections (10 µm) were cut such that each section was perpendicular to the urothelial surface. Sections were then thawed onto chrome-alum-gelatine-coated slides and processed for immunocytochemistry.
Sections were dried for 60 min at room temperature followed by three washes with Tris-buffered saline (TBS, pH 7.6), and thereafter incubated overnight with primary antibodies at 4 °C. To visualize PGP9.5 we used rabbit anti-PGP9.5 (1:2000; Abd Serotec, Oxford, UK); the selectivity and an estimate of the detection limit of these antibodies was reported previously [24–26]. The mouse antibody against vimentin (Sigma-Aldrich) was diluted 1:5000. The goat antibody against the M3 receptor (Santa Cruz Laboratories, CA, USA) was diluted 1:300. Pre-absorption of the anti-M3 antibody (1:300) was done by incubating the antibody with or without 10 µg/mL of the peptide against which the antibody was raised. Thereafter the antibody solution or antibody plus peptide solution was applied to the sections.
After overnight incubation with the primary antibodies diluted in TBS containing 0.3% (v/v) Triton X-100 (TBS-T), sections were washed in TBS, TBS-T and TBS, each wash step lasting 15 min. Rabbit primary antibody was visualized using Alexa Fluor 488 donkey antirabbit IgG (H + L) conjugate (Molecular Probes, Carlsbad, CA, USA), diluted 1:100 in TBS-T. Mouse primary antibodies were visualized with Alexa Fluor 488 donkey antimouse IgG conjugate (Molecular Probes), diluted 1:100. Goat primary antibodies were visualized with Alexa Fluor 594 donkey antigoat IgG conjugate (Molecular Probes). Sections were incubated with the secondary antibodies for 60 min at room temperature in the dark. After washing with TBS-T and TBS, sections were mounted with TBS-glycerol.
In all, five bladders were used in the study. Typically, for each bladder and for each antibody combination, staining was done in duplicate and repeated on at least two separate days. Observations were accumulated from the different slides and from the different bladders.
Sections were analysed and photographed using a light microscope with a ×4, ×10, ×20 and ×40 objective. For the detection of Alexa 488 fluorescence we used a narrow band-pass MNIBA-filter and for the detection of Alexa 594 we used a filter with a narrow excitation band, the U-M41007A filter (both filters are from Chroma Technologies, Rockingham, VT, USA). The microscope was equipped with a cooled CCD digital video camera. Images were stored digitally in 16-bit using the computer program analySIS® Vers.3.0. (Soft Imaging System, Münster, Germany). The number of grey values was reduced using a linear function into 4095. Images were arranged with the program Adobe Photoshop 5.5 or 7.0.1 (San Jose, CA, USA).
For Western blots of transfected M3 cells, HEK-293 cells were grown with Dulbecco’s modified Eagle medium, supplemented with 2 mm l-glutamine, 10% (v/v) fetal bovine serum and penicillin (100 U/mL)/streptomycin sulphate (0.1 mg/mL), at 37 °C in a humidified 5% CO2 environment. Transfections were performed for 15 h using ProFection Mammalian Transfection System-Calcium Phosphate (Promega, Madison, WI, USA), following the manufacturer’s recommendations with the ORF Clone of Homo sapiens cholinergic receptor M3 (CHRM3, OriGene, Rockville, MD, USA).
To obtain cell extracts, growing cultures were rinsed with ice-cold PBS and homogenized on ice bed with 25 mm Tris-HCl pH 7.5, 150 mm NaCl, 0.5% Triton X-100, 1 mm phenylmethylsulphonyl fluoride and 10 µg/mL leupeptin. Mixtures were cleared by centrifugation at 500g for 10 min, the protein concentration determined, and stored at −70 °C. Protein was measured using BSA dilutions (4, 2, 1, 0.5, 0.25, 0.125, 0.0625 and 0 mg/mL) with the Biorad protein measurement device (Bio-Rad Laboratories, Inc., Hertfordshire, CA, USA) according to the manufacturer’s instructions.
Western blot analyses were done under reducing conditions following standard procedures and using the Odyssey infrared imaging system (Li-cor Biosciences, Lincoln, Nebraska, USA). Primary antibodies used for Western blotting were: goat anti-M3 antibody (Santa-Cruz, SC7474) used in a dilution of 1:500, and mouse antirabbit glyceraldehyde phosphate dehydrogenase (GAPDH, RDI, 5G4–6C5) used in a dilution of 1:3.000 000, 13063). Secondary antibodies used were: donkey anti-goat IRDye 800 (Odyssey, 926–32214), and donkey antimouse IRDye 700 (Rockland, 610-730-124) both in a dilution of 1:10 000.
To find an M3 antibody suitable for immunohistochemical studies we assessed several M3 antibodies. Figure 1 illustrates the characterization of an antibody to the M3 receptor used in the present study (Santa Cruz, goat). Panel A shows a section of guinea-pig bladder stained with this antibody. The image shows that a strong immunofluorescence signal was restricted to cells lying immediately below the urothelium and that there was faint staining in some of the umbrella cells. Staining was also seen associated with the smooth muscle (not shown). One approach to assess whether this staining was specific is to interfere with the reactive site of the antibody by pre-incubation with the peptide to which it was raised. Panel B shows that all of the staining was removed with the pre-absorbed antibody. This is one indication that the antibody binding is specific for the epitope to which the antibody was raised. A second approach was to show that the antibody reacts specifically with the M3 receptor by Western blot studies. This was done using homogenates from cells which do not express the M3 receptor (control cells) and from cells which were transfected with DNA encoding for the M3 receptor (transfected cells, see above). Panel C shows a Western blot in which the antibody detected a protein band of ≈100 kDa, the correct molecular weight for the M3 receptor. This band was only found in the transfected cells. Therefore, based on these criteria we identified an antibody that can be reliably used to explore the distribution of M3 receptors in the guinea pig bladder.
M3 receptors were located on the dense population of suburothelial ICs (Su-ICs) that lie immediately below the urothelium (Fig. 2). These cells express vimentin, identifying them as being of mesenchymal origin. The diffuse network of vimentin-positive (vim+) cells which lie within the bulk of the LP (LP-ICs) did not express M3 receptors, indicating at least two distinct types of IC in the LP. The staining associated with the Su-ICs appeared to be concentrated in the cell bodies and was particularly intense in the region of the nucleus. However, staining was also clearly associated with cell processes (Fig. 2B,C). Notably, the staining appeared to be punctate, with intense aggregations of fluorescence appearing throughout the cells. Figure 1A shows that there were a few umbrella cells that stained weakly with the M3 antibody; as shown again in Fig. 3. Here there also appeared to be a punctate distribution of M3 receptors. Not all umbrella cells stained with the M3 antibody, indicating a possible heterogeneity within these epithelial cells.
Interestingly, the ICs expressing the M3 receptors were more prominent in the lateral wall than in the bladder base and urethra (Fig. 4). In the lateral wall the Su-IC layer could be up to 5–8 cells deep, with the vast majority of the cells expressing the M3 receptor. Towards the bladder base this layer thinned and was only 1–3 cells deep, and was completely absent in the region of the bladder urethral junction (Fig. 4A,B, respectively). Comparing panels B and C it is also apparent that there appeared to be more clusters of M3 receptors in the cells of the lateral wall than in the base. This differential expression might suggest a localized regulation of receptor expression. Close examination of these different regions revealed the presence of Su-ICs (vim+) that did not express M3 receptors (M3-negative). Few such cells were seen in the lateral wall, while they were more abundant at the base (Fig. 5). In the bladder urethral junction vim+ cells were found in a suburothelial layer but these cells did not have M3 receptors.
The M3 positive Su-ICs were not directly associated with nerve fibres. Nerves were seen to run within the M3 positive Su-IC layer but the density was low. However, in the region of the bladder/urethral junction, where there are numerous nerve fibres, the M3-positive Su-ICs were absent.
In the present study we used antibodies which detect the M3 type muscarinic receptors in the guinea pig urinary bladder wall. It cannot be excluded completely that the antibody has some nonspecific binding [22,23] and that it might also recognize other muscarinic receptor isoforms. By transfecting HEK cells with the M3 receptor and showing that this receptor is actually present using the two antibodies, in combination with the detection of M3 receptor reactivity in Western blots of whole guinea-pig bladder homogenates, and the complete absence of immunostaining in the tissue after pre-absorption of the antibodies with the respective peptides, we met the criteria necessary to characterize polyclonal antibodies. However, as antibodies are affinity reagents there will always be room for discussion, as there are no pharmacological/physiological model systems available to test the M3 antibodies as, e.g. have been used to characterize the cGMP antibody . However, there is concordance between our immunohistochemical results and the physiological data already available and presented in the introduction of this report, lending additional support to our viewpoint that by using these antibodies it is possible to visualize the M3 receptor. Therefore, the discussion is presented on the basis that the antibody identifies cells expressing muscarinic receptors.
The urothelium has been shown to release prostaglandin [28–31], ATP , nitric oxide (NO) [32–34] and acetylcholine [11,35] primarily in response to stretch. This release is thought to be the initial step in a system involved in detecting bladder volume. As the bladder fills the urothelium is stretched, and substances are released that result in the activation and modulation of afferent nerves. There is experimental evidence showing the modulation of afferent nerve firing by ATP  and indirect evidence for the involvement of NO  and acetylcholine  in the modulation of the afferent limb of the micturition reflex. Prostaglandin, ATP, acetylcholine and NO appear to be released throughout the bladder, in the lateral wall and dome. In these regions the density of afferent nerves is low. Therefore, is seems likely that substances released in these regions subserve functions other than direct neuro-activation or neuromodulation.
The possibility has been considered that there is more indirect involvement of substances released from the urothelium and afferent nerves [6,39]. For example, it was suggested that there are specialized cells in the LP, i.e. myofibroblasts, that lie close to afferent nerve fibres . Cells purported to be these myofibroblasts respond to ATP . This has led to the idea that the myofibroblasts contract in response to ATP, so distorting and activating the adjacent afferent nerves. Although an interesting concept, there is no direct experimental evidence for the operation of such an arrangement in the bladder wall. Based on the original description of myofibroblasts in the LP is seems unlikely that these distributed cells are those described as Su-ICs. It is more likely that the myofibroblasts lie within the regions of LP-ICs. A further aspect, suggesting that myofibroblasts and Su-ICs are different, is that the Su-IC layer in the lateral wall is poorly innervated with sensory fibres (Fig. 6) .
The absence of nerves associated with the Su-IC layer further suggests that these cells have no efferent supply and so are not under the influence of any significant neural control. If the M3 receptors are functional and activated by acetylcholine then a key question is the source of the acetylcholine. It has been shown that acetylcholine is released from the urothelium in response to stretch , and thus this is a potential source. The data from these human studies suggests that the outer cells of the urothelium express the enzyme choline acetyltransferase and are thus responsible for the synthesis of acetylcholine. A similar observation, locating choline acetyltransferase to the umbrella cells, has been made in guinea pig urothelium (de Vente and Gillespie, unpublished observations). Thus, the likely elements of this signalling system involve the release of acetylcholine from the urothelium and the activation of the underlying suburothelial cells.
The network of vim+ Su-ICs extends from the bladder base, over the lateral wall and into the dome. The co-expression with the M3 receptor is primarily located in the lateral wall where the cell layer is several cells thick. This region must therefore represent the region of the bladder where this mechanism is most active and most relevant. Towards the bladder urethral junction this system disappears, the expression of the M3 receptor declines and the proportion cells which are only vim+ is increased. The M3 mechanism is therefore absent in the region of the bladder urethral junction.
The cells which lie in the suburothelium of the bladder were first described by Smet et al. as a population of cells that responded to exogenous NO with an increase in cGMP. The physiological role of these cells and the significance of this responsiveness were not considered. Subsequent work has confirmed the sensitivity of this cell layer to NO and shown that it lies close to cells in the basal urothelium that express neuronal NO synthase, the enzyme responsible for generating NO . This suggested that there is a transfer of signals between the urothelium and Su-ICs.
It was shown that, in the guinea pig, the basal and intermediate cells of the urothelium express the enzyme cyclooxygenase I . Therefore, these cells are likely to be responsible for generating prostaglandin from the urothelium. The cellular target for this prostaglandin is unknown but there are preliminary data suggesting that the Su-ICs express the type 2 prostaglandin receptor (EP2) . In other tissues the EP2 receptor is linked to adenylate-cyclase and, when activated, generates an increase in intracellular cAMP. Thus, a second urothelial derived signal, prostaglandin, has the potential to interact with the Su-ICs and to generate a second cascade of intracellular signals.
The present data show that these Su-ICs also express M3 receptors, suggesting that they can respond to acetylcholine. M3 receptors belong to the subset of muscarinic receptors that couple to G proteins, and when activated result in an increase in diacylglycerol and inositol-triphosphate. The inositol-triphosphate generated then initiates the release of intracellular calcium. This series of events might occur within the Su-ICs.
It would thus appear that the Su-ICs receive multiple inputs via the urothelial-derived signals. These signals are likely to be involved in activating or modulating different pathways within the Su-ICs. These systems might act synergistically to produce or modulate a specific response. Alternatively, they might be antagonistic and, depending upon their level, integrate to produce a specific response. The Su-ICs could therefore be a point of integration of signals derived from the urothelium. The signals generated by the urothelium and the receptors and pathways they might act upon associated with the Su-ICs are illustrated in Fig. 7. The output of the Su-ICs and the physiological system that they might control or regulate are unknown.
It has been known for many years that the urothelial cell layer of the bladder expresses M2 and M3 receptors, with amount of M2 being greater than M3. This conclusion is based on studies of the relative expression of the appropriate mRNA and from studies of radio-ligand binding [18,48]. The techniques used cannot identify the specific cell types expressing these receptors, and the question of the location of these receptors has remained open for some time. Although the present observations give no information on the location of the M2 receptors they do indicate the Su-ICs as the location of the M3.
In conclusion, the characterization of an antibody which is specific for the M3 receptor has identified a novel signalling system in the bladder wall. The suggestion that the Su-ICs represent a cell type that can integrate the plethora of urothelial signals is an additional novel concept. These ideas now pose intriguing questions about the generation and modulation of the urothelial signals, how these signals interact and influence Su-ICs, and finally what systems are activated following Su-IC activation. Overall, the major question must be to discover the physiological system that these processes underpin.
We are very grateful to the BJU Int for their generous support. This work was supported by a BJU Int Collaborative Research Award to G.A. van Koeveringe and J.I. Gillespie.