Renal Physiology

Transient receptor potential vanilloid type 1 channels act as mechanoreceptors and cause substance P release and sensory activation in rat kidneys

Nan-Hsiung Feng, Hsang-Hsing Lee, Jeng-Chaun Shiang, Ming-Chieh Ma


Stimulation of capsaicin receptors results in an increase in afferent renal nerve activity (ARNA), but it is unclear how capsaicin contributes to sensory activation intrarenally. Here, we studied the relationships between capsaicin receptor activation, substance P (SP) release, and the sensory response in the rat renal pelvis. Immunoblots showed that one of the capsaicin receptors, transient receptor potential vanilloid type 1 channel (TRPV1), was found in various renal tissues and was especially abundant in the renal pelvis, where most sensory nerve fibers originate. Interestingly, immunolabeling showed colocalization of TRPV1, SP, and the panneuronal marker PGP9.5 in the renal pelvis. Electrophysiological recordings showed that SP and capsaicin activated the same mechanosensitive ARNA in a single-unit preparation. Intrapelvic administration of capsaicin or a specific TRPV1 agonist, resiniferatoxin, resulted in a dose-dependent increase in multi-unit ARNA and SP release, and these effects were blocked by the TRVP1 blocker capsazepine. Inhibition of the SP receptor by L-703,606 largely prevented capsaicin- or resiniferatoxin-induced ARNA. Capsazepine also prevented intrapelvic pressure (IPP)-dependent ARNA activation and contralateral diuresis/natriuresis in the renorenal reflex at an IPP of 20 mmHg, but had no effect at an IPP of 50 mmHg. These data indicate that TRPV1, a low-pressure baroreceptor, is present in the renal pelvis and exclusively regulates neuropeptide release from primary renal afferent C-fibers in response to mechanostimulation.

  • neurokinin-1 receptor
  • renorenal reflex
  • capsaicin receptor
  • intrapelvic pressure

capsaicin-sensitive afferent nerves containing substance P (SP) and calcitonin gene-related peptide (CGRP) that trigger renal sensory responses are mainly located in the renal pelvis (12, 1620, 2426). Activation of these afferent nerve fibers by mechanostimulation evokes an inhibitory renorenal reflex by withdrawing efferent renal sympathetic nerve activity (ERSNA) and increasing urine output (20, 2426). We previously showed that, when the extracellular fluid in rats is expanded by intravenous infusion of 150 mM NaCl, renal afferent nerve activity (ARNA) and intrapelvic pressure increase simultaneously and that this is associated with a decrease in ERSNA (2426). This indicates a contributory role of renal afferent nerve activation to the maintenance of body fluid balance. Several mediators, such as prostaglandin, bradykinin, angiotensin, and nitric oxide, have been shown to be involved in regulation of renal pelvic SP release and ARNA activation (1820), but the mechanisms for transducing mechano-stimuli and for the subsequent SP release are unclear. Intriguingly, defects in a regulator of SP release or the SP system itself, resulting in a decreased renal sensory response, have been proposed in different rat models of kidney diseases associated with body fluid imbalance (7, 18, 2426).

The transient receptor potential vanilloid type 1 channel (TRPV1) is also referred to as the capsaicin receptor (13). TRPV1 expression is seen in a high percentage of primary afferent neurons that project to cardiovascular and renal tissues (38). Treatment of TRPV1-expressing cells with capsaicin, resiniferatoxin (RTX), noxious heat, or a low pH generates an inward, but nonselective, cation current (5, 37). In mice lacking TRPV1, the sensitivity to noxious heat and low pH is largely attenuated, suggesting that TRPV1 is essential for pain sensation (4). Immunostaining studies have shown that TRPV1-expressing neurons in the rat urinary tract contain SP or CGRP (2, 3). Direct stimulation of TRPV1s in primary sensory nerve endings by capsaicin results in calcium influx-dependent neuropeptide release (1, 10). However, the function of TRPVs, especially that of the mechanosensitive TRPV1, in regulating the ARNA response in the kidney is unknown.

Stimulation of capsaicin receptors results in an increase in ARNA (17), but it is unclear how capsaicin contributes to renal sensory activation. Moreover, Zhu et al. (41) showed that activation of intrapelvic TRPV1s by capsaicin induces natriuresis and diuresis. We therefore asked whether TRPV1s in the renal pelvis might directly regulate SP release, which has an effect on ARNA activation and the subsequent renorenal reflex.



Female Wistar rats, weighing 200–220 g, were used. All animal experiments and care were performed in accordance with the Guide for the Care and Use of Laboratory Animals (published by National Academy Press, Washington, DC, 1996). All the protocols used in this study were approved by the Laboratory Animal Care Committee of Fu-Jen Catholic University.

Detection of TRPV1 in renal tissues.

Rats were anesthetized with intraperitoneal pentobarbital sodium (60 mg/kg) and then perfused transcardiacally with 0.1 M PBS (pH 7.4) as described previously (14). Both kidneys were rapidly dissected and one postfixed in 4% wt/vol paraformaldehyde in PBS and the other stored at −80°C for protein expression analysis.

Western blotting was performed as described previously (2426). Plasma membrane and cytosolic protein fractions of renal cortex, medulla, and pelvis were prepared using a commercial extraction kit (BioVision) and subjected to electrophoresis. Blotting was performed on rabbit antibodies against TRPV1 (Santa Cruz Biotechnology; 1:1,000 dilution), neurokinin-1 (NK-1) receptor (Santa Cruz Biotechnology; 1:500 dilution), or SP (Biomeda; 1:1,000 dilution), or mouse antibody against Na+-K+-ATPase (Affinity Bioreagents; 1:2,000 dilution), or actin (Biomeda; 1:2,000 dilution). After being washed, the membrane was incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG antibody (Vector Laboratory, Burlingame, CA) and bound antibody was visualized using a commercial ECL kit (Amersham Bioscience) and Kodak film. The densities of the bands for TRPV1, NK-1 receptor, Na+-K+-ATPase, SP, and actin, with respective molecular masses of ∼100, 79, 110, 16, and 40 kDa, were determined using an image analyzing system (Diagnostic Instruments).

The postfixative kidneys were stored in 10% sucrose in 4% paraformaldehyde solution at 4°C, then embedded in O.C.T. compound (Tissue-Tek, Sakura Finetek, Torrence, CA), and frozen at −20°C until used to prepare 5-μm sections on a cryostat (Microm, Heidelberg, Germany) which were thaw-mounted on coated slides. After rehydration and being washed with PBS, the sections were processed for indirect immunofluorescence, using antibodies from the same sources as above plus anti-PGP9.5 antibody (Chemicon, Temecula, CA). After being blocked with 5% skim milk in PBS for 1 h at room temperature (RT), the sections were incubated overnight at 4°C with rabbit anti-TRPV1 antibody diluted 500-fold in 5% normal rabbit serum in PBS and then for 1 h at RT with rhodamine-Red-X-conjugated donkey anti-rabbit antibody (Jackson ImmunoResearch, West Grove, PA; 1:100 in 5% skim milk). After detection of TRPV1, the tissue sections were incubated overnight at 4°C with mouse anti-SP antibody (1:400 in PBS) or anti-PGP9.5 antibody (1:500 in PBS) and then for 1 h at RT with FITC-conjugated donkey anti-mouse antibody (Jackson ImmunoResearch; 1:200 in 5% skim milk), and then examined on a Olympus BX51 microscope (Tokyo, Japan) equipped with a fluorescent image analytic system (Diagnostic Instruments) at ×400 magnification. Nuclei were counterstained using DAPI. The specificity of each antibody was tested by preincubating it for 4 h at 4°C with the specific blocking peptide provided by Santa Cruz Biotechnology (150 μg/ml) before carrying out the test.

General surgical preparation for the in vivo study.

Rats were anesthetized with pentobarbital sodium (60 mg/kg ip) and the trachea, external jugular vein, and carotid artery were cannulated for, respectively, spontaneous ventilation, continuous saline infusion at 1.2 ml/h, and measurement of the mean arterial blood pressure (MABP), as described previously (2426). The left kidney was exposed via a left flank incision and the ureter was cannulated near the pelvis with a combined PE-50/10 catheter for perfusion with saline (150 mM NaCl) or drugs.

Renal pelvic perfusion.

Two PE-10 catheters with a heat-pulled tip (<20 μm) were placed inside a 5-cm length of PE-50 catheter in the left ureter and extending 1 to 2 mm beyond the tip of the PE-50 catheter as described previously (2426). The tips of the catheters were placed together in the left ureter near the renal pelvis, allowing the renal pelvis to be perfused with saline at 20 μl/min via the two PE-10 catheters; this perfusion did not affect renal pelvic pressure or ARNA (see Fig. 2A). The effluent was drained away by the PE-50 catheter, with a dead space of <4.5 μl. The other end of the 5-cm PE-50 catheter was connected to a T-tube connector to record the intrapelvic pressure (IPP) change. The third end of the T-tube connector was connected to a 60-cm length of PE-50 tubing which could be raised to increase the IPP.

Recording of ARNA.

The techniques for recording single- or multi-unit renal nerve activity have been described previously (2426). For single-unit recording, one nerve bundle was repeatedly split with fine forceps until the nerve spikes showed similar amplitudes, durations, and shape on the oscilloscope (Tektronix). Nerve activity was amplified and filtered using an AC preamplifier (NL104a Digitimer, Hertfordshire, UK) and filter (NL125, Digitimer), and the amplified signals were selected using a spike trigger (NL201, Digitimer) and counted on a pulse integrator (NL601, Digitimer). Neural activity was transformed into spike counts, integrated over 10-s intervals, and displayed continuously using MP150 AcqKnowledge software (Biopac). The proximal part of the nerve fibers was transected to record single-unit ARNA. For multi-unit recording, an identical preparation to the above was used to record ARNA from an intact nerve bundle. Nerve activities were allowed to stabilize for 1.5 h after preparation. Single- or multi-unit recordings of ARNA were identified using an increase in IPP, which does not act as a specific stimulus for other subtypes of renal sensory receptor.

Effects of TRPV1 agonist and blocker on ARNA.

The protocol consisted of basal, experimental, and recovery (Rec) periods of 10, 3, and 10 min, respectively (n = 12). Stock solutions of capsazepine and capsaicin were prepared in ethanol and diluted in saline, the final ethanol concentration being <0.1%. Saline or capsazepine (Sigma, 5 μg/ml) was continuously infused at 10 μl/min via one PE-10 catheter during the basal and experimental periods. The other PE-10 catheter was used for saline infusion during the basal period and for infusion of various concentrations of capsaicin (0.02, 0.06, or 0.2 μg/ml) or RTX (Sigma, 0.5, 1.5, 2.5, or 5.0 μg/ml in saline; all infusions at 10 μl/min) during the experimental period. During the recovery period, saline was infused through both PE-10 catheters. A 3-min experimental period was used, as a previous study showed that long-term capsaicin treatment destroys renal sensory nerves (9).

The renal pelvic effluent during RTX application was collected for analysis of SP release.

Effects of SP and capsaicin on single-unit ARNA.

The functional coexistence of the NK-1 receptor and TRPV1 in the same mechanosensitive nerve terminal was tested using single-unit recording of ARNA. Renal pelvic mechanostimulation was performed by raising the 60-cm PE-50 catheter to increase the IPP by 20 mmHg for 3 min as described above. SP (Sigma, 10 μg/ml, dissolved in saline) or capsaicin (0.2 μg/ml) was then administered for 3 min at 10 μl/min via one PE-10 catheter, whereas the other PE-10 catheter was continuously perfused with saline at 10 μl/min. Each test was bracketed by a 10-min basal period and a 10-min recovery period of saline perfusion.

Effect of NK-1 receptor inhibition on capsaicin-induced ARNA.

After perfusion with saline for 10 min, capsaicin (0.06 μg/ml) or RTX (2.5 μg/ml) was infused into the renal pelvis for 3 min to activate ARNA and then washed out by saline in a 10-min recovery period. The specific nonpeptide NK-1 receptor antagonist, L-703,606 (Sigma, 0.1 mg/ml) (27), was then infused via one PE-10 catheter for 10 min before 3-min treatment with the same dose of capsaicin or RTX as before via the other infusion catheter (n = 12). L-703,606 was prepared in ethanol and diluted in saline, with a final ethanol concentration of <0.1%. After testing the effect of NK-1 receptor inhibition, the ARNA in response to capsaicin or RTX was again tested in the absence of L-703,606 to confirm the effect of the blocker was not due to rundown of ARNA (data not shown).

Effects of capsazepine on the renorenal reflex.

The IPP was increased by 10, 15, 20, or 50 mmHg for 3 min by raising the 60-cm length of PE-50 tube. Saline or capsazepine (5 μg/ml) was infused into the renal pelvis as above for 5 min before mechanostimulation and throughout the IPP increase. At the IPP of 20 or 50 mmHg, contralateral (right kidney) urine samples were collected in addition to recording ARNA and MABP to evaluate the renorenal reflex response in the presence or absence of capsazepine (n = 12).

SP assay.

Tissues sampled from the renal cortex, medulla, and pelvis were homogenized at 4°C in the presence of the endopeptidase inhibitor thiorphan (Sigma; final concentration 10 μmol/l) to minimize SP degradation (4, 6, 8). Thiorphan was also added to renal pelvic effluents immediately after collection. SP was measured using an enzyme-linked immunoassay, as described previously (26).

Data treatment and chemical analyses.

Systemic hemodynamics and renal excretion were measured and averaged over each period. For multi-unit ARNA, postmortem nerve activity was subtracted from all values of renal nerve activity. The ARNA responses to the various stimuli were averaged using AcqKnowledge software and the baseline value was subtracted from the treatment (drug testing) or recovery period values as the change in ARNA (ΔARNA). The ΔARNA were then calculated as a percentage of the baseline value, i.e., ΔARNA (%).

In the contralateral kidney, the urine volume was determined gravimetrically, the urinary sodium concentration was measured using an electrolyte analyzer (Dri-Chem 3500i, Fuji, Tokyo, Japan), and the urinary flow rate (UV) and urinary sodium excretory rate (UNaV) were expressed per gram of kidney weight.

Statistical analysis.

Numerical data are presented as means ± SE. Differences between groups were analyzed using an unpaired t-test or one-way ANOVA, with a posttest using Duncan's multiple-range test. Differences were regarded as significant at P < 0.05.


Localization of TRPV1 in renal tissue.

The location of TRPV1 in the rat kidney was determined by immunofluorescence staining (Fig. 1A). In both the renal cortex and medulla, TRPV1 was not homogenously distributed, showing strong expression in the apical membrane of the tubular lumen and lower expression in the tubular cells (Fig. 1A, top left and middle). Most TRPV1s were found in the distal tubules and collecting ducts. TRPV1 was also seen in the tubular cells. In the renal pelvic wall (Fig. 1A, top right), TRPV1 was mainly expressed in the fibers between uroepithelial and smooth muscle cells; some also being seen in the uroepithelium. The negative control, in which the anti-TRPV1 antibody was preincubated with a specific blocking peptide, gave no signal (Fig. 1A, bottom). Doublelabeling of kidney sections for TRPV1 and SP showed that both were found in most fiber bundles in the renal pelvis (Fig. 1B). Furthermore, Fig. 1C shows that TRPV1 expressed in the renal pelvic wall colocalized with the panneuronal marker PGP9.5. Interestingly, PGP9.5 was also mainly found between uroepithelial and smooth muscle cell layers; some also being seen beneath smooth muscle cells in the renal pelvis. These results strongly suggest the presence of TRPV1 in sensory nerve structures of the renal pelvis.

Fig. 1.

Transient receptor potential vanilloid type 1 channel (TRPV1) expression in the rat kidney. A: TRPV1 immunoreactivity (red color indicated by arrowheads) in the renal cortex, medulla, and pelvic wall. Bottom: pictures show that preabsorption of the antibody with specific blocking peptide blocks TRPV1 labeling. G, glomerulus. Horizontal bars = 10 μm. Nuclei were counterstained using DAPI (blue color). B and C: immunofluorescence doublelabeling for TRPV1 (top) and substance P (SP) or PGP9.5 (middle) in the renal pelvic wall showing colocalization in most nerve terminals (indicated by arrowheads in the merged pictures; bottom). D and E: typical Western blots showing expression of TRPV1 in plasma membrane and cytosolic fractions (D) and expressions of neurokinin (NK)-1 receptor, TRPV1, and SP in plasma membrane and cytosolic fractions (E) of renal cortex (C), medulla (M), and pelvis (P). Na+-K+-ATPase or actin was used as loading control. Statistical results for the ratio of the density units (DU) for TRPV1 compared with Na+-K+-ATPase and actin (n = 4 in D) and for the ratio of the DU for TRPV1 and NK-1 receptor compared with Na+-K+-ATPase and for SP compared with actin (n = 6 in E). *Significant difference (P < 0.05) compared with the corresponding tissue in D or compared with renal cortex in E.

Using Western blot analysis, Fig. 1D shows the presence of TRPV1 in the plasma membrane and cytosolic fraction. However, cytosolic expression of TRPV1 was significantly less than membrane expression in all tissue parts.

Previous studies showed that NK-1 receptors are expressed in the renal pelvis, most being present in the plasma membrane (6, 26). It was therefore of interest to know whether TRPV1 was present in the same membrane preparation. As shown in Fig. 1E, Western blots showed that the NK-1 receptor was found in the plasma membrane fraction in various parts of the kidney, being most abundant in the renal pelvis (left). Interestingly, TRPV1 was also found in the same fraction and was most highly expressed in the renal pelvis. In the renal cortex and medulla, TRPV1 was expressed, respectively, at 49 ± 5 and 71 ± 9% of the levels in the renal pelvis (Fig. 1E, middle). Using immunohistochemical staining, Liu and Barajas (22) showed that most renal sensory neurons in the renal pelvic wall contain SP. In the present study, Fig. 1D, right, showed that SP was found in the cytosolic fraction and its distribution was similar to that of TRPV1. This is consistent with results for SP measured by enzyme-linked immunoassay showing that the highest amount of SP was present in the renal pelvis (9.4 ± 0.8 pg/mg of protein compared with 2.8 ± 0.7 in renal medulla and 1.6 ± 0.3 in renal cortex).

SP and capsaicin activate the same mechanosensory nerve.

We then examined whether the potent TRPV1 agonist capsaicin acted on the same SP receptor-containing mechanosensory nerve fibers. A typical spike chosen for recording is shown in Fig. 2B. The estimated conduction velocity, calculated as the spike duration divided by the distance between electrodes, from 6 single-unit recordings was 2.04 ± 0.13 m/s, suggesting that the nerve recorded was a C-fiber. As shown in a typical single-unit ARNA recording (Fig. 2C) and Table 1, nerve firing and the transformed spike count were increased by an IPP of 20 mmHg. After recovery, SP treatment (10 μg/ml) resulted in increased nerve discharge in the same preparation (Fig. 2D), as did subsequent treatment with capsaicin (0.2 μg/ml; Fig. 2E). IPP or drug treatment had no effect on arterial pressure (AP).

Fig. 2.

Activation of single-unit renal mechanosensory nerve. A: original traces showing that renal pelvic perfusion does not affect the arterial blood pressure (AP), intrapelvic pressure (IPP), afferent renal nerve activity (ARNA) discharges, and integrated counts of the ARNA. B: typical ARNA spike in the single-unit ARNA recording is shown to explain the method for estimating the conduction velocity of renal afferent nerve. Effects of an increase in IPP (C) or intrapelvic infusion of SP (D) or capsaicin (E) on the AP, ARNA discharges, and integrated single-unit ARNA recorded simultaneously.

View this table:
Table 1.

Responses of single-unit ARNA to IPP, SP, and capsaicin

Capsazepine attenuates the TRPV1-mediated ARNA increase and SP release.

We then asked whether TRPV1 activation had any effect on ARNA and SP release. In the typical recordings shown in Fig. 3A, capsaicin administered via the intrapelvic route caused a dose-dependent increase in the ARNA. The AP and IPP were unaffected by capsaicin, but a slight decrease in the AP and a slight increase in the IPP (from 3.2 to 5.7 mmHg) were seen at the concentration of 0.2 μg/ml. Figure 3B shows that capsaicin caused an increase in the ΔARNA and that this effect was abolished by pretreatment with the TRPV1 antagonist capsazepine (5 μg/ml). Capsazepine alone had little effect on the basal ΔARNA.

Fig. 3.

Renal sensory response and SP release in relation to TRPV1 activation or inhibition. A: original traces showing the effect of intrapelvic infusion of capsaicin on the AP, IPP, and integrated multi-unit ARNA in one rat. Top and bottom: ARNA traces show, respectively, 5 s of recording during the basal period and 5 s during the peak response after capsaicin treatment. The horizontal bar indicates infusion of 0.02, 0.06, or 0.2 μg/ml of capsaicin. B and C: statistical results for the percentage change in the ΔARNA in response to capsaicin (B) or resiniferatoxin (RTX; C) with or without capsazepine pretreatment (n = 12 for each test). D: SP release in the renal pelvis measured in the presence of RTX with or without capsazepine pretreatment (n = 12 for each test). *Significant difference (P < 0.05) compared with the saline-treated period. †Significant difference (P < 0.05) between the 2 treatments at the same dose of agonist.

Application of a highly specific TRPV1 agonist, RTX, caused a dose-dependent increase in ΔARNA and this effect was also completely blocked by capsazepine pretreatment (Fig. 3C).

Analysis of the effluent collected during RTX treatment showed an increase in SP release which was largely prevented by capsazepine pretreatment (Fig. 3D).

NK-1 receptor blockade inhibits the TRPV1-mediated ARNA response.

The typical multi-unit ARNA traces in Fig. 4A demonstrate that intrapelvic application of the nonpeptide NK-1 receptor antagonist L-703,606 at 0.1 mg/ml blocked both the capsaicin (left)- and RTX (right)-induced increase in ARNA in the same cell, but had no effect on the AP or IPP. The summarized results (Fig. 4B) show that 0.06 μg/ml of capsaicin or 2.5 μg/ml of RTX resulted in a ΔARNA of 122.4 ± 22.5 and 98.8 ± 17.2%, respectively, and that both responses were largely prevented by L-703,606 pretreatment (ΔARNA of 20.1 ± 15.9 and 10.2 ± 12.7%, respectively).

Fig. 4.

Effect of an NK-1 receptor inhibitor on ARNA responses to TRPV1 activation. Representative traces (A) and statistical results (B) showing the effect of 0.2 μg/ml of capsaicin or 2.5 μg/ml of RTX on ARNA with or without NK-1 receptor inhibition by infusion of L-703,606 (n = 12). *Significant difference (P < 0.05) compared with baseline. †Significant difference (P < 0.05) compared with capsaicin or RTX treatment in the absence of L-703,606.

Capsazepine decreases the mechanostimulation-induced reflex response.

Figure 5A shows typical recordings in response to mechanostimulation by increasing the IPP in the presence (right) or absence (left) of capsazepine. The summarized results (Fig. 5B) demonstrate that, at IPPs of 10, 15, and 20 mmHg in the absence of capsazepine, the ΔARNA increased in a pressure-dependent manner without any effect on the AP and that the effect was partially prevented by intrapelvic administration of capsazepine.

Fig. 5.

Renal sensory and reflex responses to renal pelvic mechanoreceptor activation after TRPV1 inhibition. A: typical traces showing the effect of mechanoreceptor activation caused by increasing the IPP to 10, 15, or 20 mmHg, on the AP and integrated multi-unit ARNA in the absence (left) or presence (right) of capsazepine (Capz). B: statistical results for the ARNA responses in the presence or absence of Capz (n = 12 for each test). C: renorenal reflex induced by increasing the IPP to ∼20 mmHg in the absence (left) or presence (right) of Capz (n = 12 for each test). The mean arterial blood pressure (MABP), urinary flow rate (UV), and urinary sodium excretory rate (UNaV) were recorded in the contralateral right kidney, while the ARNA was recorded in the left kidney. Rec, recovery period. *Significant difference (P < 0.05) compared with the basal state. †Significant difference (P < 0.05) compared with treatment in the absence of Capz (vehicle).

Figure 5C shows that stimulation of the renal pelvic mechanoreceptor elicited an inhibitory renorenal reflex. Capsazepine alone had no effect on the MABP. An IPP increase to ∼20 mmHg in the left kidney resulted in an increase in the ΔARNA (from 3.6 ± 15.8 to 198.1 ± 43.8%), accompanied by increases in the UV (from 4.2 ± 0.8 to 7.4 ± 1.5 μl·min−1·g−1) and UNaV (from 0.09 ± 0.01 to 0.14 ± 0.02 μmol·min−1·g−1) in the right kidney (left). In contrast, after capsazepine treatment (Fig.5C, right), no contralateral diuresis/natriuresis was seen with the same IPP increase.

Effect of capsazepine at a high IPP.

Strong mechanostimulation caused by increasing the IPP to ∼50 mmHg elicited a large ARNA increase associated with reflex hypotension, and capsazepine pretreatment reduced the effect slightly, but nonsignificantly (Fig. 6, A and B). This again suggests that the attenuated ARNA response after capsazepine treatment in response to gradual increases in IPP in the range of 10–20 mmHg was not due to desensitization of nerve activity. Both the UV and UNaV were increased in the contralateral kidney by raising the IPP, and capsazepine pretreatment had no effect (Fig. 6C), even at higher concentrations (10 and 20 μg/ml; data not shown).

Fig. 6.

Effect of TRPV1 inhibition on high IPP-induced renal sensory and reflex responses. A: typical traces showing the effect of mechanostimulation by increasing the IPP to ∼50 mmHg on the AP and integrated multi-unit ARNA in the absence (top) or presence (bottom) of Capz. B: statistical results for the ARNA responses in the presence or absence of Capz (n = 12 for each test). C: renorenal reflex response in the left kidney induced by increasing the IPP in the absence (left) or presence (right) of Capz (n = 12 for each test). *Significant difference (P < 0.05) compared with the basal state.


As shown schematically in Fig. 7, this study presents functional evidence for the presence of TRPV1 in the renal pelvis and suggests that it is the candidate mechanoreceptor for renal C-fiber activation. It is interesting that the effect of TRPV1 on the sensory response was dependent on the intrapelvic neuropeptide system. These results are consistent with previous observations suggesting that SP is a candidate neuropeptide for mechanoactivation (12, 1620, 2426).

Fig. 7.

Scheme showing how TRPV1 interacts with the NK-1 receptor to affect ARNA generation. On mechanostimulation, TRPV1 in the sensory nerve terminals allows calcium entry, thus increasing SP release into the pelvic space where it acts on the NK-1 receptor to generate ARNA.

The TRPV1 was unevenly distributed in the rat kidney (Fig. 1). Less TRPV1 expression was seen in the intracellular compartment in all renal tissues. Previous studies indicated that TRPV1 is expressed not only in the plasma membrane but also in the endoplasmic reticulum, where it functions as a Ca2+ release channel (16, 23, 29). However, a 10-fold higher concentration of capsaicin is needed to stimulate the cytosolic TRPV1 than the plasma membrane TRPV1 (39). Moreover, subcellular translocation of TRPV1 has been suggested to contribute to the potentiation or sensitization of receptor function in abnormal pain sensation (34, 42). Whether the above mechanism for TRPV1 sensitization is involved in regulation of SP release and the renal sensory response requires further study. Western blots showed that it was most abundant in the renal pelvis, an area in which the NK-1 receptor and SP were also expressed. Outside the central nervous system (CNS), TRPV1 is suggested to be located in the peripheral end of primary sensory nerve and, in addition to responding to thermal change and decreased pH, also activates the sensory response in a different way by enhancing the release of neurotransmitters or neuromodulators (15, 28). In the rat urinary tract, primary sensory fibers expressing TRPV1 are extremely abundant in the epithelial layer or closely apposed to smooth muscle cells and are colocalized with SP or CGRP (2, 10). Consistent with this, our results showed that TRPV1-immunoreactive fibers were located in the nerve structure (evidenced by PGP9.5 staining) and just beneath or among the uroepithelial cells of the renal pelvic wall and were colocalized with SP and contributed to its release (Figs. 1B and 3D). SP and capsaicin activated the same mechanosensory nerve, providing further evidence for the functional coexistence of TRPV1s and NK-1 receptors (Fig. 2). Moreover, the single-unit ARNA was not affected by intrapelvic administration of 900 mM NaCl, brief renal arterial occlusion-induced ischemia, or venous occlusion (data not shown), indicating a specific response of the ureteropelvic type of mechanoreceptor MRu, but not other renal receptors (7, 11).

Previous studies suggested that TRPVs are found not only in nerve but also in various tubular segments (8, 13, 21). For example, TRPV4 in mammal kidneys has been suggested to be an osmo-sensor. In trpv4knockout mice, Liedtke and Friedman (21) found defective osmoregulation in the CNS when faced with hyperosmotic challenge. Whether TRPV1 plays the same role as other medullary TRPVs requires further study. Little TRPV1 immunoreactivity was seen in the renal cortex in the present study in agreement with a previous observation that sensory nerves are extremely rare in the renal cortex (9). In terms of channel composition, several splice variants of TRPV1, such as stretch-inactivated channel, 5′ sv, and TRPVVAR, have been identified in the kidney (32, 35, 36). These variants may serve to modulate canonical TRPV1s or directly allow calcium influx in response to RTX. Although we did not identify the type of renal pelvic TRPV1, these variants have been suggested to be associated with mechanosensitivity in different tissues (28).

Since SP release is calcium dependent, it is rational to deduce that TRPV1 may directly increase SP release from nerve terminals by a mechanism involving calcium influx (4). Our results support a direct effect of TRPV1 on SP release. Scotland et al. (33) showed that, in cardiovascular tissue, an increase in the luminal pressure in the mesenteric arteries causes release of 20-hydroxyeicosatetraenoic acid, which, in turn, activates TRPV1s expressed on C-fibers, leading to vasoactive neuropeptide release. As it can regulate cation influx, activation of TRPV itself might directly generate ARNA; however, our results using an NK-1 receptor antagonist proved that TRPV1-induced sensory activation was dependent on NK-1 receptor function, which is downstream of TRPV1.

In pressure transduction in the renal pelvis, TRPV1 is indispensable for the mechanoactivation of renal afferent nerves. The important function of TRPV1 as an initiator of renal sensory response also has a profound effect on the whole reflex response, as TRPV1 blockade totally abrogated contralateral diuresis/natriuresis in the renorenal reflex (Fig. 5). Removal of these sensory fibers in neonatal rats by a large dose of capsaicin or RTX is suggested to impair the diuretic, natriuretic, and kaliuretic responses to saline or water load (27, 31). Recently, an interesting study showed that perfusion of capsaicin into one side of the renal pelvis causes diuresis and natriuresis bilaterally and this is blocked by capsazepine antagonism of TRPV1 channel function (41), indicating the functional presence of TRPV1 in the renal pelvis. Moreover, the TPRV1-mediated excretory response is related to renal nerve function, as the increased excretory response disappears after acute renal denervation (41). In the present study, we showed anatomic and electrophysiological evidence that TRPV1 is present in the sensory nerve structure inside the renal pelvis and contributes to SP release in ARNA activation, triggering the inhibitory renorenal reflex and causing the diuretic and natriuretic response. TRPV1 has been suggested to act as a baroreceptor in the carotid sinus (40). In bladders excised from trpv1-null mice, purinergic signaling in response to stretch is greatly diminished (3). TRPV1s transduce a broad range of pressure (from 10 to 100 mmHg) in vivo (2, 3, 30, 33, 42). In the present study, TRPV1 contributed to ARNA activation in the IPP range of 10–20 mmHg. However, at a high IPP (50 mmHg), TRPV1 inhibition did not abrogate ARNA activation and the reflex response (Fig. 6), indicating that the attenuated ARNA responses seen after TRPV1 inhibition in the low IPP range were not due to desensitization of nerve activity. The intrapelvic TRPV1 probably acts as a low-pressure baroreceptor and there is probably an alternative sensory transmission for a high IPP.


This work was supported by grants from the National Science Council of the Republic of China (NSC95-2320-B-030-006-MY2 to M.-C. Ma) and from the Military Kaohsiung General Hospital (9507 to M.-C. Ma, N.-H. Feng, and J.-C. Shiang).


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract