Differential Localizations of the Transient Receptor Potential Channels TRPV4 and TRPV1 in the Mouse Urinary Bladder

Abstract

We studied the localization and physiological functions of the transient receptor potential (TRP) channels TRPV1 (TRP vanilloid 1) and TRPV4 (TRP vanilloid 4) in the mouse bladder, because both channels are thought to be mechanosensors for bladder distention. RT-PCR specifically amplified TRPV4 transcripts from the urothelial cells, whereas TRPV1 transcripts were barely detectable. ISH experiments showed that TRPV4 transcripts were abundantly expressed in the urothelium, whereas TRPV1 transcripts were not detectable in the urothelial cells. Immunoblotting and IHC studies showed that TRPV4 proteins were mainly localized at the basal plasma membrane domains of the basal urothelial cells. In contrast, TRPV1-immunoreactivities were found not in the urothelial cells but in the nerve fibers that innervate the urinary bladder. In Ca²⁺-imaging experiments, 4α-phorbol 12,13-didecanoate, a TRPV4 agonist, and hypotonic stimuli induced significant increases in intracellular calcium ion concentration ([Ca²⁺]_i) in isolated urothelial cells, whereas capsaicin, a TRPV1 agonist, showed no marked effect on the cells. These findings raise the possibility that, in mouse urothelial cells, TRPV4 may contribute to the detection of increases in intravesical pressure related to the micturition reflex. (J Histochem Cytochem 57:277–287, 2009)

The urinary bladder urothelium can release various chemical mediators such as NO, ATP, and acetylcholine in response to thermal, mechanical, and chemical stimuli, probably allowing reciprocal communication with neighboring urothelial cells as well as nerves or other cells in the bladder wall (Birder 2005). When the intravesical pressure of urinary bladder or the degree of urothelium distention increases, ATP is released from urothelial cells (Ferguson et al. 1997) and activates purinergic receptors expressed in nearby nerve terminals within the urothelium, carrying the information to the central nervous system (Birder 2005). According to previous physiological experiments, TRPV4, a Ca²⁺-permeable stretch-activated cation channel, is expressed in rat and mouse urothelial cells (Birder et al. 2007; Gevaert et al. 2007). The activation of TRPV4 by hypotonic stimuli or 4α-phorbol 12,13-didecanoate (4α-PDD), a TRPV4 agonist, induces significant increases in [Ca²⁺]_i in rat urothelial cells, leading to ATP release (Birder et al. 2007). Stretch-induced ATP release related to the activation of TRPV4 is observed in isolated mouse bladders, and TRPV4-deficient mice exhibit abnormal frequencies of voiding and non-voiding contractions in cystometric experiments (Gevaert et al. 2007). TRPV4 is thus likely to be one of important urothelial mechanosensors for bladder distention.

TRPV1 has also been implicated in normal bladder function and is another putative mechanosensor for urothelium distention. The TRPV1 channel is expressed not only in the peripheral neurons that innervate the bladder but also in urothelial cells in rats (Birder et al. 2001). Similar to the effect of hypotonicity, the application of capsaicin, a TRPV1 agonist, triggers ATP release from cultured mouse urothelial cells (Birder et al. 2002). However, no direct evidence of mechanogating of TRPV1 in heterologous expression systems has been provided, and besides, capsazepine, a TRPV1 antagonist, had no effect on the bladder reflex activity of normal mouse bladders (Dinis et al. 2004). Therefore, the role of TRPV1 in normal bladder function is still controversial.

Thus far, information on the expression of TRPV1 in mouse urothelia at the molecular level is not available, and the histological relationship between TRPV1 and TRPV4 in the mouse bladder has not been clearly described. To address the above discrepancies, we studied the expression of TRPV1 and TRPV4 in the mouse bladder, using a combination of molecular biological, morphological, and physiological approaches.

Materials and Methods

The Center of Experimental Animal Sciences at Nagoya City University gave us permission for the following experiments.

Isolation of Urothelial Cells

The preparation of urothelial cells was carried out as previously described with some modifications (Birder et al. 1998; Truschel et al. 1999; Birder et al. 2001). Briefly, bladders excised from anesthetized C57BL/6J and TRPV1 knockout (TRPV1^−/−) mice (allele symbol TRPV1^tm1Jul; Jackson Labs, Bar Harbor, ME) (Caterina et al. 2000) at 8 weeks of age were cut open and gently stretched (urothelial side up). The muscle layers were dissected away after incubation with 2.5 mg/ml dispase (Invitrogen; Carlsbad, CA) for 2 hr at room temperature. Subsequently, the urothelium was treated with 0.05% trypsin–0.02% EDTA for 20 min at 37C. After being centrifuged, collected urothelial cells were resuspended in keratinocyte media (Invitrogen). The single cell suspension produced (100,000–150,000 cells/ml) was subsequently used for RT-PCR and Ca²⁺-imaging experiments. It was confirmed that almost all cells isolated in this culture system were cytokeratin 7 (a mammalian urothelial marker) positive (data not shown).

Isolation of Dorsal Root Ganglion Cells

L4–L6 dorsal root ganglia (DRGs) were quickly excised bilaterally from anesthetized C57BL/6J mice (8 weeks) and incubated in GIBCO-BRL Neurobasal-A Media (Invitrogen) containing 2 mg/ml collagenase, 1 mg/ml trypsin, and 3 mM CaCl₂ at 37C for 45 min. After centrifuged, DRG cells were resuspended in Neurobasal-A media supplemented with 5% B27 supplement, 0.25% l-glutamine (200 mM), and 1% penicillin-streptomycin (all from Invitrogen), plated on poly-d-lysine–coated glass coverslips, and incubated at 37C in a 95% air–5% CO₂ atmosphere saturated with water vapor. Subsequently, Ca²⁺-imaging experiments were carried out 2–4 hr after the cell isolation.

RT-PCR

Three μg of total RNA isolated from the urothelial cells or L4–L6 DRGs was subjected to random-primed reverse transcription using SuperScript II (Invitrogen). Next, 1/40th of the sample was amplified by PCR for 35 cycles with the following primers: for the N-terminal region of TRPV4 (product length = 337 bp); 5′-ATTCCTTGTTCGACTACGGCACT-3′ (sense) and 5′-GCAATGTCCAGCAACACCGGGAT-3′ (antisense); for the C-terminal region (329 bp), 5′-TGGCGAGGTCATCACGCTCTTCACA-3′ (sense) and 5′-GGTACACAAGCAGGAAGCGGAAGA-3′ (antisense); for the N-terminal region of TRPV1 (466 bp), 5′-CTATGATCGCAGGAGCATCTTCGA-3′ (sense) and 5′-GAACTTCACAATGGCCAGCTGGTT-3′ (antisense); for the C-terminal region (511 bp), 5′-TGCCTGTGGAGTCCCCACCACACAA-3′ (sense) and 5′-AGGGAGAAGCTCAGGGTGCGCTTGA-3′ (antisense); for β-actin (298 bp), 5′-GATCCTGACCGAGCGTGGCTACA-3′ (sense) and 5′-TGAAGTGTGACGTTGACATCCGT-3′ (antisense). As a negative control, the total RNAs isolated were processed without reverse transcription. The PCR products obtained were separated by 1% agarose gel electrophoresis. The molecular identity and homogeneity of the resultant PCR products were checked by DNA sequencing (data not shown).

ISH

C57BL/6J mice (8 weeks) were decapitated under deep anesthesia (Nembutal, IP, 50 mg/kg), and their bladders and L4–L6 DRGs were quickly removed. Histological preparations (10 μm thick) and ISH were performed essentially as described elsewhere (Ueda et al. 2003a; Ugawa et al. 2006). [³⁵S]UTP-labeled cRNA probes were generated from nucleotide sequences as follows: for TRPV4, nucleotides 530–1726 (GenBank accession no. NM022017); for TRPV1, nucleotides 1564–2486 (GenBank accession no. NM001001445).

Immunoblotting

Whole bladders excised from anesthetized C57BL/6J or TRPV1^−/− mice and HEK293T cells transfected with empty or mouse TRPV4-containing pcDNA3.1 expression vectors (Invitrogen) were homogenized in 50 mM Tris-HCl, pH 7.5/5 mM EDTA/150 mM NaCl/1% Triton X-100/complete protease inhibitors (Roche Diagnostics; Basel, Switzerland). The supernatants were collected after centrifugation, and their total protein concentrations were measured. Equal amounts of protein (25 μg) were resolved by SDS-PAGE (8.5%). TRPV4 was detected using a commercially available anti-TRPV4 antibody raised against the synthetic peptide CDGHQQGYAPKWRAEDAPL (amino acid residues 853-871 of rat TRPV4) (Alomone; Jerusalem, Israel). This was followed by a horseradish peroxidase–labeled anti-rabbit IgG secondary antibody, and it was visualized using the ECL Western Blotting Detection kit (GE Healthcare; Buckinghamshire, UK). Nonspecific reactions were assessed by preabsorption treatment with 10⁻⁶ M of the antigen peptide. TRPV1 protein expression in the mouse whole bladders was also examined using two commercially available anti-TRPV1 antibodies P-19 (near N-terminal) or R-18 (C-terminal; Santa Cruz Biotechnology, Santa Cruz, CA).

Fluorescent IHC

Fresh frozen bladder and L4–L6 DRGs sections were isolated from 8-week C57BL/6J or TRPV1^−/− mice using a cryostat after decapitation under deep anesthesia. After air drying, the sections were fixed in 4% paraformaldehyde/0.1 M phosphate buffer (pH 7.4; 4% PFA) for 20 min at 4C and incubated with the anti-TRPV4 (1:1000) and/or anti-cytokeratin 7 (1:500) antibodies (Dako; Carpinteria, CA) in PBS containing 0.3% Triton X-100 overnight at 4C. Some sections were fixed in 4% PFA for 20 min at 4C and incubated with the anti-TRPV1 antibody P-19 or R-18 (1:500) and an anti-PGP (protein gene product) 9.5 antibody (Santa Cruz Biotechnology) (1:500). These were followed by an Alexa 488–conjugated donkey anti-rabbit IgG (Invitrogen), an Alexa 488–conjugated donkey anti-goat IgG, an Alexa 594–conjugated donkey anti-goat IgG, or a rhodamine-conjugated donkey anti-guinea pig IgG (AP193R; Chemicon, Temecula, CA).

Immunogold Electron Microscopy

C57BL/6J mouse bladders (8 weeks) were prefixed in 4% paraformaldehyde/0.1 M PB for 20 min at 4C, washed with PB, and incubated with the anti-TRPV4 antibody diluted 1:1000 in PB containing 0.15% Tween 20 (PB/T) overnight at 4C. After washing, the samples were incubated with 5-nm gold anti-rabbit antibody (British BioCell International; South Glamorgan, UK) in PB/T at 4C overnight and postfixed in 2.5% glutaraldehyde for 1 hr. Subsequently, they were fixed in 2% osmium tetroxide for 2 hr, dehydrated, and embedded in Epon resin, and 90-nm sections were cut, placed on copper grids, stained with uranyl acetate and lead citrate, and observed with an electron microscope JEM-2010 (JEOL; Tokyo, Japan).

Measurement of Intracellular Calcium Concentration ([Ca²⁺]_i)

Single cell suspensions of the urothelial cells were plated on pieces of coverglass using Cell-Tak (BD Biosciences; San Jose, CA) as soon as the cells were resuspended in keratinocyte medium and incubated for 30 min at 37C in a 95% air–5% CO₂ atmosphere saturated with water vapor. The fixed urothelial cells or DRG cells were incubated with a fluorescent Ca²⁺ indicator (fura-2 acetoxymethyl ester, 10 μM; Invitrogen) in assay buffer (10 mM HEPES, 130 mM NaCl, 10 mM glucose, 5 mM KCl, 2 mM CaCl₂, and 1.2 mM MgCl₂, pH 7.4, 290 mOsm) for 30 min at room temperature. The loading solution was washed out, and the cells were stimulated with capsaicin (0.1–10 μM), capsazepine (CPZ; 5 μM), 4α-PDD (10 μM), ruthenium red (RuR; 10 μM), hypotonicity (220 mOsm), or ionomycin (3 μM) using a bath perfusion system at a flow rate of 2–3 ml/min. We recorded [Ca²⁺]_i changes using an Olympus IX-70 microscope equipped with the ARGUS/HiSCA system (Hamamatsu Photonics; Hamamatsu, Japan). Acquisitions and analyses of the fluorescence images were performed with the ARGUS/HiSCA version 1.65 software (Ueda et al. 2003b). We checked the condition of the cells using calcium ionophore (Ionomycin; 3 μM) and found that ∼80% of the isolated cells were viable (data not shown).

Results

Expression of TRPV1 and TRPV4 mRNAs in Mouse Urothelial Cells

We carefully isolated the total RNA from urothelial cells of mouse urinary bladders and performed RT-PCR. As shown in Figure 1, a couple of products with the predicted sizes and nucleotide sequences of the N-terminal (exons 2–4) and C-terminal (exons 9–12) ends of TRPV4 mRNA were amplified from the cells (Lanes 1 and 3, respectively). On the other hand, PCR products corresponding to the N-terminal (exons 2-5) and C-terminal (exons 12–14) ends of TRPV1 were barely detected in the urothelial samples prepared (Lanes 5 and 7, respectively), whereas both transcripts were intensely amplified from the DRG (Lanes 9 and 11, respectively) where the TRPV1 channels are primarily located (Tominaga et al. 1998). N- and C-terminal ends of TRPV4 transcripts were also detected in the DRG (Lanes 13 and 15, respectively). Control β-actin fragments amplified in parallel are shown in the bottom panel. These results suggest that the expression level of TRPV1 transcripts is much lower than that of TRPV4 transcripts and raise the possibility that functional TRPV1 channels hardly exist in the mouse urothelial cells.

Figure 1.

RT-PCR analysis of transient receptor potential channels 4 and 1 (TRPV4 and TRPV1) transcripts in mouse bladder urothelia and dorsal root ganglia. N- and C- indicate N-terminal and C-terminal PCR products of each channel, respectively. RT(–) means templates without reverse transcriptase. Control β-actin fragments are shown in the bottom panel. The size of the DNA standards is indicated on the left.

To further evaluate these findings, we next performed ISH experiments on the urinary bladder sections. It was found that TRPV4 transcripts were strongly expressed in the urothelial cells (Figure 2A). At a higher magnification, the signals were likely to be mainly located in the intermediate and basal cells of the urothelia (Figure 2B). Labeling was not detected when the corresponding sense probe was used (Figure 2C). The expression of TRPV4 transcripts was also examined in mouse DRG neurons, and faint hybridization signals were sparsely distributed on the sections compared with sense controls (Figures 2D and 2E, respectively).

Figure 2.

Expression of TRPV4 and TRPV1 transcripts in the mouse urothelium and dorsal root ganglion (DRG) neurons by ISH. (A,B) Strong expression of TRPV4 transcripts was observed in the mouse urothelia at low (A) and high (B) magnifications when the TRPV4 antisense probe was used. (C) No signal was observed when the TRPV4 sense probe was used. (D) In the DRG sections, faint expression of TRPV4 transcripts was observed using the TRPV4 antisense probe. (E) The TRPV4 sense probe did not detect any signals on the DRG sections. (F,G) Specific hybridization signals were undetectable in the urothelia at low (F) and high (G) magnifications when the TRPV1 antisense probe was used. (H) The control bladder section hybridized with the TRPV1 sense probe. (I) Strong expression of TRPV1 transcripts in the DRG neurons. (J) The control DRG section hybridized with the TRPV1 sense probe. Bar = 50 μm.

Despite the detection of TRPV1 transcripts by RT-PCR, no specific signal for TRPV1 was observed in the urothelial cells (Figures 2F and 2G). (Bladder sections hybridized with the corresponding sense probe are shown in Figure 2H.) In control experiments, strong expression of TRPV1 transcripts was detected using the same antisense probe in the DRG sections (Figure 2I), whereas no signal was found in the DRG with the sense probe (Figure 2J). These results, together with the RT-PCR data, showed that TRPV4 transcripts were definitely expressed in the urothelial cells and that the total amount of TRPV1 transcripts in the cells was not enough for the detection by the ISH method.

Expressions of TRPV4 Proteins in Mouse Urothelial Cells

To confirm the expression of TRPV4 in the urothelial cells at the protein level, we performed immunoblotting analyses using a rabbit polyclonal antibody against the C-terminal region of TRPV4 (Figure 3A). As shown in Lane 1, the antibody recognized a single band of ∼110 kDa in the total extracts of mouse TRPV4-expressing HEK293T cells. This immunostaining pattern was not found in the total extracts of HEK293T cells transfected with the vector alone (Lane 2), showing that the antibody recognized mouse TRPV4 protein. Next, the same experiments were carried out with lysates derived from the mouse urinary bladders (Figure 3A). It showed that a single prominent band with a similar molecular mass (∼110 kDa) was detected in the lysates (Lane 4). Preabsorption treatment with 10⁻⁶ M antigen peptide completely prevented the labeling (Lane 3). These results showed the specificity of the TRPV4 antibody and the expression of TRPV4 proteins in the mouse urinary bladder.

Figure 3.

Immunoblotting and IHC analyses of TRPV4 in the mouse urothelium. (A) Detection of TRPV4 protein (arrowhead) in total extracts of TRPV4-expressing or non-expressing HEK293T cells (Lanes 1 and 2, respectively), and detection of TRPV4 protein in the bladder extracts in the presence or absence of 10⁻⁶ M antigen peptide (Lanes 3 and 4, respectively). Labeled bands of actin are shown below as positive controls. Mr, molecular mass standards. (B) Cellular localization of TRPV4 protein in the mouse urothelium at a low magnification. (Left) Cytokeratin 7 as an urothelial marker. (Middle) TRPV4-positive urothelial cells. (Right) A merged image of the left and middle panels. (C) The TRPV4 immunoreactivity was most prominent at the basal membranes of the basal urothelial cells. (Left) Cytokeratin 7. (Middle) TRPV4-positive urothelial cells. (Right) a merged image. (D) Nomarski image of C. (E) TRPV4 immunoreactivities in the urothelia after preabsorption treatment (left). (Right panel) Nomarski image of the left panel. Bar = 50 μm.

We subsequently performed IHC analyses. It was found that the antibody against the intermediate filament, cytokeratin 7, which is known to be expressed in mammalian urothelia (Erman et al. 2006), stained all the layers of the urothelium tested (Figure 3B, left panel). TRPV4 immunoreactivities were only found within the urothelia in the same sections (middle panel; merged image, right panel). At a higher magnification (Figures 3C and 3D), the TRPV4 immunoreactivity was the most prominent at the basal membranes of the basal urothelial cells (Figure 3C, middle and right panels). Preabsorption treatment with specific antigen peptide (10⁻⁶ M) for TRPV4 completely inhibited the staining (Figure 3E). These findings are similar to the TRPV4 localization found in the rat urothelium (Gevaert et al. 2007).

Immunoelectron Microscopy

To assess the ultrastructural localization of the TRPV4 channel in the mouse bladder, we used immunoelectron microscopy. Because the antibody was raised against the intracellular domain of the channel, cell permeabilization with detergent was unavoidable in the immunostaining, leading to substantial disruption of the plasma membranes. However, we succeeded in detecting the immunoreactivity in the bladder by minimizing the detergent concentration. It was found that gold particles, which represented the intracellular epitopes recognized by the TRPV4 polyclonal antibody, were sparsely located around the basal urothelial cell membranes just adjacent to the loose connective tissue (Figure 4A). At a higher magnification, the immunogold labelings were likely to be observed intracellularly around the basal plasma membrane domains (Figure 4B). In other instances, the epitopes were likewise observed around the intracellular side of the basal plasma membrane of the basal urothelial cells (Figures 4C and 4D). Preabsorption of the antibody with the antigen peptide (10⁻⁶ M) confirmed the specificity of the data (Figure 4E). We sought to find the gold particles in the apical plasma membrane domains of basal urothelial cells and in the superficial and intermediate parts of the urothelia, but could not detect any signals, indicating that the TRPV4 channels are mainly located in the basal plasma membrane domains of basal urothelial cells in the mouse.

Figure 4.

Immunoelectron microscopic analyses of TRPV4 in the mouse urothelium. (A) Gold particles, which represented the intracellular epitopes recognized by the TRPV4 polyclonal antibody, were sparsely and intracellularly located near the basal urothelial cell membranes just adjacent to the loose connective tissue (arrow and arrowhead). (B) Gold particles indicated by the arrow in A at a higher magnification. (C,D) In other instances, the epitopes are observed around the intracellular side of the basal plasma membrane domains of the basal urothelial cells. (E) Preabsorption of the antibody with the antigen peptide (10⁻⁶ M) confirmed the specificity of the data. N, nucleus. Bar = 200 nm.

Expressions of TRPV1 Proteins in Nerve Terminals Innervating Mouse Bladders

Using two commercially available antibodies raised against the near N-terminal region (P-19) and C-terminal region (R-18) of the TRPV1 channel, immunoblotting analyses were performed with lysates extracted from wild-type (TRPV1^+/+) and TRPV1^−/− mouse bladders, which contained the urothelia, smooth muscle layers, and nerve endings innervating the bladders (Figure 5A). It showed that both antibodies recognized several proteins with various molecular masses in the TRPV1^+/+ lysates (Lanes 1 and 3). However, a single band with a molecular mass of ∼100 kDa, which is almost the same as the predicted molecular mass of the mouse TRPV1 channel (Bodó et al. 2005), was absent in the TRPV1^−/− lysates, whereas other protein stainings were still observed (Lanes 2 and 4). These results implied that, in the mouse bladder, the antibodies were able to react with TRPV1, and that the single ∼100-kDa protein band was probably caused by the TRPV1 protein.

Figure 5.

Immunoblotting and IHC analyses of TRPV1 in the mouse urothelium. (A) Detection of TRPV1 protein in the total extracts of wild-type mouse bladders, using commercially available P-19 and R-18 anti-TRPV1 antibodies. The single bands with a molecular mass of ∼100 kDa (arrowhead; Lanes 1 and 3) were specifically absent in the TRPV1^−/− bladder lysates (Lanes 2 and 4). Mr, molecular mass standards. KO, TRPV1^−/−. (B) IHC localization of TRPV1 protein in wild-type mouse urothelia shown by P-19. (Left) Protein gene product (PGP) 9.5 as a nerve-ending marker. (Middle) TRPV1 immunoreactivities in the urothelium. (Right) A merged image of the left and middle panels. (C) Localization of TRPV1 protein in the TRPV1^−/− mouse urothelium examined using P-19. (Left) PGP 9.5–positive neurons in the urothelium. (Right) Absence of TRPV1 immunoreactivities. (D) TRPV1 immunoreactivities in the wild-type DRG neurons as a positive control of P-19. (E) The TRPV1 immunoreactivities found in D were undetectable in the TRPV1^−/− DRG sections (top). (Bottom) A Nomarski image of the top panel. (F) PGP 9.5 (red, left) and TRPV1 (green, middle) double positive neurons (yellow, right) were found in the wild-type mouse urothelium using R-18. (G) The TRPV1 immunoreactivities found in F were undetectable in the TRPV1^−/− bladder sections. Bar = 50 μm.

Although the expression level of TRPV1 mRNA seemed to be relatively low, substantial TRPV1 channel proteins may be localized in the mouse urothelium. We performed TRPV1 immunostaining of TRPV1^+/+ and TRPV1^−/− mouse bladder sections using those anti-TRPV1 antibodies. In the TRPV1^+/+ mice, nerve fibers were stained by a well-known pan-neuronal marker PGP 9.5 (Figure 5B, left panel). When P-19 was used, strong immunofluorescent signals were concentrated in the nerve fibers innervating the urothelia, and faint immunofluorescence was also detected in the urothelial cells (Figure 5B, middle panel; merged image, right panel). In the TRPV1^−/− mice, PGP 9.5–positive nerve endings were not stained with P-19, although the fluorescence in the urothelial cells still existed in substantial quantities (compare left and right in Figure 5C), suggesting that the P-19 immunoreactivity in the TRPV1^+/+ nerve endings was probably caused by TRPV1 proteins and that TRPV1 was not reflected in the immunofluorescence observed in the urothelial cells. The P-19 antibody was also functional on mouse DRG sections, because strong immunoreactivities were present in a subset of the TRPV1^+/+ DRG neurons (Figure 5D) and absent in the TRPV1^−/− DRG cells (Figure 5E). We further carried out the same experiments using another TRPV1 antibody, R-18, and similar results were obtained as shown in Figures 5F and 5G. Taken together, these IHC analyses suggested that the mouse urothelial cells rarely or never exhibited the TRPV1 channel.

Functional Analyses of TRPV1 and TRPV4 in Mouse Urothelial Cells

To confirm the presence of the TRPV4 channel and scarce distribution of functional TRPV1 in mouse urothelial cells, we physiologically studied the effects of several agonists of both channels on freshly isolated mouse urothelial cells. We used primary cultured urothelial cells immediately after isolation to avoid possible artificial upregulations of proteins caused by enzymatic treatment, although previous studies have used the cells at least 48 hr after isolation (Birder et al. 1998; Truschel et al. 1999). As shown in Figure 6A, hypotonic solutions (220 mOsm) evoked marked increases in [Ca²⁺]_i in isolated urothelial cells, and these responses were markedly diminished by 10 μM ruthenium red (RuR), a non-selective antagonist of the TRPV cation channel family. Similar [Ca²⁺]_i responses were observed in urothelial cells isolated from TRPV1^−/− mice (Figure 6B), raising the possibility that TRPV4 (and not TRPV1), at least in part, contributes to the hypotonicity-induced [Ca²⁺]_i increases. We tested a total of 361 urothelial cells from the TRPV1^+/+ mice and 82 urothelial cells from the TRPV1^−/− mice, and ∼39.3% (142/361) and 35.4% (29/82) of the cells responded to the hypotonic stimuli, respectively. The ionomycin (3 μM) application showed that ∼80% of each isolated cells were viable in this experimental system, which means that ∼49% and 44% of the isolated cells from the TRPV1^+/+ and TRPV1^−/− mice responded to the stimuli, respectively.

Figure 6.

Intracellular Ca²⁺ ([Ca²⁺]_i) response in primary cultured mouse urothelial cells and DRG neurons. (A,B) Hypotonic stimuli (220 mOsm) evoked marked [Ca²⁺]_i increases that were abolished by extracellular ruthenium red (RuR, 10 μM) in the urothelial cells isolated from TRPV1^+/+ (A) and TRPV1^−/− (B) mice. (C) TRPV4 agonist 4α-phorbol 12,13-didecanoate (4α-PDD; 10 μM) evoked [Ca²⁺]_i increases that were partially blocked by extracellular RuR. Note that the cell did not respond to 1 μM capsaicin (Cap). (D) Capsaicin (500 nM) evoked [Ca²⁺]_i increases in primary cultured DRG neurons, and the response was almost completely inhibited by capsazepine (CPZ, 5 μM). A calcium ionophore (Ionomycin, 3 μM) was applied as a positive control at the end of the experiments.

In other urothelial cells isolated from TRPV1^+/+ mice, 10 μM 4α-PDD, a TRPV4 agonist, caused significant [Ca²⁺]_i increases, which were substantially reduced in the presence of 10 μM RuR (Figure 6C). In contrast, capsaicin (up to 10 μM) showed no apparent effect on the cells, which was in good agreement with our morphological data. Approximately 27.2% of the cells examined responded to the 4α-PDD treatment (69/254), estimating that ∼34% of the viable urothelial cells were responsive to the treatment. Capsaicin at drug concentrations ranging from 100 nM to 10 μM evoked no [Ca²⁺]_i increases in 514 urothelial cells isolated from TRPV1^+/+ mice. To confirm that the capsaicin used in our Ca²⁺-imaging system was functional, we isolated mouse DRG neurons in parallel and applied up to 500 nM of capsaicin on the cells. As shown in Figure 6D, the extracellular capsaicin apparently caused [Ca²⁺]_i increases, which were almost completely blocked by 5 μM capsazepine (CPZ), a TRPV1-selective antagonist. Based on these findings, we concluded that almost no functional TRPV1 channels were expressed in the urothelial cells tested.

Discussion

TRPV4 is a member of the heat-activated Ca²⁺-permeable TRPV cation channel family (Guler et al. 2002) and is activated by mechanical stimuli such as cell swelling and/or hypo-osmolality (Liedtke et al. 2000), as well as by diverse chemical compounds including 4α-PDD (Watanabe et al. 2002). It is widely thought that urothelial cells act as transducers whereby afferent neurons, through urothelial mechanoafferent transduction, are involved in the micturition process (Birder 2005; Gevaert et al. 2007). According to earlier studies, urothelial TRPV4 is likely to be a mechanosensor, transforming mechanical stimuli (bladder distention) into ATP signals. Actually, a decreased intravesical stretch-evoked ATP release was found in isolated whole bladders from TRPV4^−/− mice (Gevaert et al. 2007).

In this study, we morphologically confirmed that TRPV4 channels were expressed in mouse urothelial cells, especially at the basal plasma membrane domains of the basal urothelial cells. In our physiological studies, the isolated urothelial cells showed marked increases in [Ca²⁺]_i when 4α-PDD and hypotonic stimuli were applied. Our results support the possible involvement of TRPV4 in the micturition reflex. It is noteworthy, however, that the proportion of urothelial cells responding to the hypotonic stimuli (∼49%) was larger than that of the 4α-PDD–sensitive cells (∼34%), suggesting the existence of other hypotonicity/cell swelling-sensitive Ca²⁺-permeable channels in mouse urothelial cells. In addition, more than one half of the freshly isolated urothelial cells could not respond to the hypotonic or 4α-PDD stimuli, which may suggest the heterogeneity of gene expression patterns among the urothelial cells.

Recently, it has been reported that the TRPV4 channel is located not only in the mouse urothelium but also in the bladder smooth muscle (Thorneloe et al. 2008). Interestingly, TRPV4 activation with a potent TRPV4 activator GSK1016790A contracted TRPV4^+/+ mouse bladders in vitro, both in the presence and absence of the urothelium; this effect was undetected in TRPV4^−/− bladders. Direct infusion of GSK1016790A into the bladders of TRPV4^+/+ mice induced bladder overactivity with no effect in TRPV4^−/− mice. In another study, the amplitude of the spontaneous contractions in explanted bladder strips from TRPV4^−/− mice was significantly reduced, and the TRPV4^−/− mice exhibited a lower frequency of voiding contractions and a higher frequency of non-voiding contractions (Gevaert et al. 2007). In vivo, the TRPV4 channels in the urothelium and in the bladder smooth muscle could cooperatively play important roles in urinary bladder function that includes not only the urothelium-mediated transduction of intravesical mechanical pressure but also an ability to contract the bladder. Apart from those TRPV4 channels, TRPV4 proteins are also located in mouse sensory neurons (Suzuki et al. 2003), which is in good agreement with our RT-PCR and ISH data. These findings suggest that the TRPV4 channels may be expressed on the bladder sensory terminals themselves and that the channels may mechanically be gated by bladder distension without any release of chemical mediators from the urothelium. The detailed localization of TRPV4 on the sensory terminals and possible contribution of the neuronal TRPV4 to the detection of bladder distension remain to be addressed.

TRPV1 has also been reported to be expressed in urothelial cells in rats (Birder et al. 2001) and could be an important mechanosensor molecule that triggers ATP release from the cells (Birder 2005). In contrast to TRPV4, however, the role of TRPV1 in the normal urothelium is not clear, because significant discrepancies remain about the bladder functions of TRPV1^−/− mice. In urothelial cells isolated from TRPV1^−/− mice, ATP release caused by hypotonic stimuli is markedly diminished, and the mutants have a higher frequency of low-amplitude, non-voiding bladder contractions than their wild-type littermates (Birder et al. 2002). According to other independent studies, the frequencies of reflex bladder contractions of TRPV1^+/+ and TRPV1^−/− mice were exactly the same (Avelino and Cruz 2006).

Under the circumstances, we re-evaluated whether functional TRPV1 is expressed in mouse urothelial cells but failed to obtain any evidence of its expression. First, both the N-terminal and C-terminal regions of TRPV1 were hardly amplified from the isolated urothelial cells by RT-PCR. Second, our ISH analyses showed that TRPV1 transcripts were undetectable in the urothelial cells. Third, faint staining by the anti-TRPV1 antibodies in the urothelial cells was not abolished in the TRPV1^−/− mice. Finally, in our physiological experiments, capsaicin failed to evoke any increases in the [Ca²⁺]_i of isolated urothelial cells, although previous Ca²⁺-imaging studies described mouse urothelial cells that reacted to capsaicin (Birder et al. 2001). We used isolated urothelial cells just after isolation in our Ca²⁺-imaging studies. On the other hand, the previous experiments used cells at least 48 hr after isolation, and, during the long incubation time, the gene expression pattern of TRPV1 was likely to be significantly altered, leading to the artificial upregulation of the protein. All these findings strongly suggest that functional TRPV1 channels may hardly exist in mouse urothelial cells in normal conditions. Several IHC studies have reported that TRPV1 is also expressed in human urothelial cells (Lazzeri et al. 2004,2005; Apostolidis et al. 2005a,b). However, other studies have cast doubt on those experiments (Dinis et al. 2005). It might be necessary to reconfirm that TRPV1 is expressed in bladder urothelial cells in rats and humans.

It has previously been reported that TRPV1 immunoreactivities are expressed in the nerve fibers that innervate the urinary bladder in rats and humans (Birder et al. 2001; Yiangou et al. 2001), and intravesical application of resinifera toxin (RTX), a TRPV1 agonist, decreased the immunoreactivity in bladder nerve fibers, possibly because of the RTX-induced desensitization and related downregulation of the TRPV1 channel (Avelino et al. 2002; Brady et al. 2004). According to our IHC data, TRPV1 immunoreactivity is found in mouse bladder nerve endings. In particular, its strong immunoreactivity is observed in the nerve fibers located in the suburothelium. These results are rather convincing because the TRPV1^−/− mice lacked a similar immunoreaction. In single-unit bladder nerve recordings, low-threshold neuronal responses were attenuated in TRPV1^−/− mice compared with the TRPV1^+/+ littermates, whereas high-threshold sensitivity was unchanged (Daly et al. 2007), suggesting that those neuronal TRPV1 channels in the suburothelium seem to be needed for the normal excitability of low-threshold bladder fibers. The neuronal TRPV1 is also likely to have pathophysiological roles, contributing to overactive bladder and pain (Avelino and Cruz 2006).

In conclusion, our data support the idea that TRPV4 functions as one of urothelial transducers for the detection of bladder distension. Molecular identities of other possible mechanosensors and their relationships with TRPV4 in vivo remain to be elucidated. TRPV1 is unlikely to be located in mouse urothelial cells and would work mainly in bladder nerve terminals. Detailed molecular mechanisms for the possible involvements of the neuronal TRPV1 channels in the micturition reflex should be further addressed.