Celia Cruz, Department of Experimental Biology, FMUP, Centre for Medical Research, Fifth floor, Rua Dr Plácido Costa, 91, 4200-450 Porto, Portugal. e-mail: firstname.lastname@example.org
What's known on the subject? and What does the study add?
The interaction between the TRPV1 and NGF systems has been addressed only in the context of acute somatic pain. The present study expands this view and indicates that this interaction remains operative and is important as a mechanism for chronic visceral pain and dysfunction. Moreover, it further stresses the need to develop more specific and effective TRPV1 antagonists for clinical use.
• To explore the role of transient receptor potential vanilloid 1 (TRPV1) in the excitatory effects of chronic administration of nerve growth factor (NGF) on bladder-generated sensory input and reflex activity.
• To explore new therapeutic targets for bladder dysfunction.
MATERIALS AND METHODS
• Wild-type (WT) and TRPV1 knockout (KO) mice received daily intraperitoneal injections of NGF (1 µg/10 g) or saline for a period of 4 days, during which time thermal sensitivity was evaluated daily. On the 5th day, mice were anaesthetized and cystometries were performed. The frequency, amplitude and area under the curve (AUC) of bladder reflex contractions were determined.
• c-Fos expression was evaluated on L6 spinal cord sections of WT and TRPV1 KO mice treated with saline or chronic NGF by immunohistochemistry.
• TrkA receptor staining intensity was determined in L6 spinal cord sections and respective dorsal root ganglia of WT and TRPV1 KO mice.
• Repeated administration of NGF induced thermal hypersensitivity in WT but not in TRPV1 KO mice.
• The frequency of bladder contractions of saline-treated WT and TRPV1 KO mice was similar, the values respectively being 0.45 ± 0.12/min and 0.46 ± 0.16/min. Treatment with NGF enhanced bladder reflex activity in WT mice to 1.23 ± 0.41/min (P < 0.05). In NGF-treated KO mice, the frequency of bladder contractions was 0.60 ± 0.05/min. Irrespective of treatment, no differences were observed in the amplitude of bladder contractions of WT and TRPV1 KO mice. The AUC was significantly increased in NGF-treated WT-mice, when compared with saline-treated WT-mice. No changes were found in AUC of saline-treated and NGF-treated TRPV1 KO mice.
• Chronic administration of NGF resulted in a significant increase of spinal c-Fos expression in WT mice (P < 0.05 vs KO animals), but not in TRPV1 KO animals.
• TrkA expression was similar in WT and TRPV1 KO mice.
• NGF-induced bladder overactivity and noxious input depend on the interaction of NGF with TRPV1.
• The lack of bladder overactivity in TRPV1 KO mice treated with NGF does not represent loss of TrkA expression.
• TRPV1 is essential for NGF-driven bladder dysfunction and represents a bottleneck target in bladder pathologies associated with NGF up-regulation.
3,3′-diaminobenzidine tetrahydrochloride as chromogen
dorsal root ganglion
Transient receptor potential vanilloid (TRPV1) is a transmembrane ion channel. It is not essential for physiological bladder function in healthy animals as TRPV1 knockout (KO) mice exhibit normal or near-normal bladder activity . In contrast, TRPV1 is essential for detrusor overactivity and increased voiding frequency accompanying acute and chronic cystitis in rodents . In overactive bladder syndrome and interstitial cystitis/bladder pain syndrome TRPV1 overexpression has been described in the bladder wall, both in urothelial cells and in suburothelial nerve fibres [2,3]. The expression of neuronal TRPV1 correlates with the intensity of bladder pain  whereas the intensity of TRPV1 in the urothelium of patients with sensory urgency inversely correlates with the volume of urine associated with the first desire to void .
Nerve growth factor (NGF) is a tissue-derived neurotrophin essential for the survival and differentiation of sensory neurons , which acts through the tyrosine kinase receptor TrkA . It is acutely released after an inflammatory insult and significantly contributes to the swift modification of the pain threshold and visceral activity . NGF has also been considered to be sufficient to elicit cutaneous hyperalgesia . Exogenous administration of NGF induces detrusor overactivity in naive rats [9,10], as well as up-regulating the expression of spinal c-Fos and the firing of bladder nociceptive fibres [9,11], in agreement with the strong expression of TrkA receptors by bladder sensory afferents .
The identification of downstream effectors of NGF-induced effects has only recently become a focus of systematic research. In this context, TRPV1 was found to be essential for the development of thermal hyperalgesia after acute administration of NGF . In mice that received intraplantar NGF, the latency of paw withdrawal to a noxious thermal stimulus, an indication of thermal hyperalgesia, was substantially decreased in wild-type (WT) mice but not in TRPV1 KO mice . In this study we evaluated if the interplay between NGF and TRPV1, important for somatic pain, is also present in NGF-mediated bladder overactivity and noxious input.
MATERIALS AND METHODS
TRPV1 KO female mice from The Jackson Laboratory (Bar Harbor, ME, USA) and WT female mice belonging to the same strain (C57BL/6; n= 6/group) from the Instituto de Biologia Molecular e Celular colony (Porto, Portugal) with an average weight of 25 g were used. Animals were maintained in the animal house at 22 °C and 60% humidity under a 12-h light/dark cycle. All experiments were carried out according to the European Commission Directive of 22 September 2010 (2010/63/EU) and the ethical guidelines for investigation of experimental pain in animals . All efforts were made to reduce the number of animals used.
The NGF was purchased from Promega (Madison, WI, USA). The antibody against TrkA receptor, made in rabbit, was purchased from Millipore, Watford, UK. The antibody against c-Fos, made in rabbit, came from Millipore, Watford, UK. Biotin-conjugated swine anti-rabbit antibody came from Dakopatts A/5 (Copenhagen, Denmark). The ABC Vectastain Elite kit (ABC, avidin–biotin complex) and the conjugate horseradish peroxidase were purchased from Vector Laboratories (Peterborough, UK). Antibodies and the ABC complex were prepared in PBS 0.1 m containing 0.3% Triton X-100 (PBST). For cystometry and terminal handling, mice received a subcutaneous bolus of urethane (1.2 g/kg) as anaesthetic.
The NGF was given as a daily intraperitoneal injection (1 µg/10 g) for 4 days. The dose of NGF was chosen according to previous studies . Sterile saline was given as a control. In addition, because NGF is known to induce thermal hyperalgesia in WT animals , the effects of repeated NGF administration on thermal sensitivity were evaluated daily (see below) to assure the biological action of NGF. On day 5, mice were anaesthetized and cystometry was performed, after which animals were perfusion-fixed and the L6 spinal cord segment was collected for c-Fos assessment.
The hot-plate test was used to measure the response latencies to thermal noxious stimuli. Hence, NGF-injected TRPV1 KO and WT mice were placed in individual chambers (10 × 20 × 14 cm) and allowed to acclimate for 5 min. Latencies were determined before and 4 h after each NGF injection. For that, animals were placed on the hot-plate apparatus (Series 8, model PE34, IITC Life Sciences, Woodland Hills, CA, USA). The platform was maintained at 35.0 ± 0.1 °C and temperature was increased up to 52.5 °C. The time spent between placement of the animal on the platform and positive behavioural responses (jumping on the platform and/or licking of the hindpaws) was registered as the response latency.
To assess bladder function, cystometry was performed in TRPV1 KO and WT mice treated with saline or NGF on day 5 (n= 6). Anaesthesia was induced by a subcutaneous injection of urethane and body temperature was maintained at 37 °C with a heating pad. The urinary bladder was exposed through an incision in the lower abdomen. A 25-gauge needle was inserted in the bladder dome and the urethra remained unobstructed throughout the recording period while saline was infused at a constant rate (1.6 mL/h). The frequency, amplitude of bladder contractions and area under the curve (AUC) were then measured for a period of 90 min. In all experiments, recordings were made after a 30-min stabilization period.
After cystometry, animals were perfused through the ascending aorta with cold oxygenated calcium-free Tyrode's solution (0.12 m NaCl, 5.4 mm KCl, 1.6 mm MgCl2.6H2O, 0.4 mm MgSO4.7H2O, 1.2 mm NaH2PO4.H2O, 5.5 mm glucose, 26.2 mm NaHCO3), followed by cold 4% paraformaldehyde. The L6 spinal cord segments were post-fixed for 4 h in the same fixative solution and cryoprotected for 24 h in 30% sucrose with 0.1% sodium azide in 0.1 m phosphate buffer. Transverse 20-µm sections from the spinal cord were cut in the freezing microtome and stored in cryoprotective solution at −20 °C until further processing. When all material was collected and cut, every second spinal section from each animal was thawed and immunoreacted against c-Fos to evaluate the expression of c-Fos. Briefly, after inhibition of endogenous peroxidase activity and thorough washes in PBS and PBST, sections were incubated in 10% normal swine serum in PBST for 2 h. Sections were then incubated for 48 h at 4 °C with a specific antibody against c-Fos (1:10 000). Subsequently, sections were washed and incubated with polyclonal swine anti-rabbit biotin-conjugated antibody (1:200). To visualize the immunoreaction, the ABC conjugated with peroxidase (1:200) method was used with 3,3′-diaminobenzidine tetrahydrochloride as chromogen (DAB; 5 min in 0.05 m Tris–HCl buffer, pH 7.4 containing 0.05% DAB and 0.003% hydrogen peroxide). Sections were mounted on gelatine-coated slides and air-dried for 12 h, cleared in xylene, mounted with Eukitt mounting medium and cover-slipped.
For analysis of TrkA expression the WT and TRPV1 KO mice (n= 4 per group) were perfusion-fixed with calcium-free Tyrode's solution followed by 4% paraformaldehyde. The L6 spinal cord segment and respective dorsal root ganglion (DRG) were collected. Transverse 20-µm sections of spinal cord sections were cut in the freezing microtome and stored in cryoprotective solution at −20 °C. Longitudinal 12-µm sections of L6 DRG were cut in the cryostat and stored at −20 °C. When all material was collected, sections were removed from the freezer and processed for TrkA immunoreaction, as described above for c-Fos expression. The antibody against TrkA was used at 1:1000 dilution.
Cystometrograms were evaluated using Data Trax software (Vs. 1.804; World Precision Instruments, Sarasota, FL, USA). The frequencies, amplitudes of bladder contractions and AUC were analysed using Kruskal–Wallis one-way repeated measures anova. Data are presented as mean value ±sd and P < 0.05 was considered significant. Statistical analysis was performed with SigmaStat 3.5 software.
The number of c-Fos immunoreactive nuclei was counted in 10 non-consecutive sections from each animal and averaged. Statistical analysis was performed using anova followed by the Student–Newman–Keuls post-hoc test, using the SigmaStat 3.5 software.
Quantification of the TrkA staining intensity was done with Fiji software (based on ImageJ, http://rsb.info.nih.gov/ij Java 1.6.0_20, 32 bit). The intensity of staining was averaged from sections per animal (DRG and spinal cord) and a reference intensity of unstained tissue was also measured by a fourth box on all sections. Background intensity was deducted from the average intensity to calculate the mean net staining intensity. The intensity of TrkA staining in WT mice treated with saline was used as control. Data were analysed by one-way anova followed by the Student–Newman–Keuls post-hoc test. The data are presented as mean value ±sd and P < 0.05 was considered significant. Statistical analysis was carried out using the Graph Pad Prism software.
Daily intraperitoneal NGF injections significantly reduced thermal latency in the hot-plate test in WT mice from the day 2 onwards (Fig. 1). The baseline temperature at which WT animals presented nocifensive behaviour (jumping on the platform and/or licking of the hindpaws) was 45.9 ± 0.9 °C and was not changed by saline treatment (Fig. 1). In contrast, in WT animals the thermal latencies after each NGF injection were significantly decreased to 45.6 ± 0.7 °C, 43.9 ± 1.1 °C, 44.8 ± 0.6 °C and 44.9 ± 0.4 °C, respectively at days 1, 2, 3 and 4, respectively (P < 0.001 vs TRPV1 KO mice at all time points; Fig. 1). In TRPV1 KO mice the baseline latency was 46.4 ± 0.6 °C and was not changed after intraperitoneal saline injections. The temperatures registered in KO mice receiving NGF were 45.5 ± 1.1 °C, 46.6 ± 0.9 °C, 46.8 ± 0.9 °C and 47.1 ± 1.3 °C at days 1, 2, 3 and 4, respectively (Fig. 1), and were not different from baseline values.
In cystometrograms obtained from saline-treated TRPV1 KO mice, we observed non-voiding small amplitude oscillations that preceded voiding contractions (Fig. 2C). These were absent in recordings from saline-treated WT animals (Fig. 2A) and were reflected in a marginally higher AUC in KO mice (Fig. 3C), although the difference did not reach significance. The frequency of voiding contractions of WT and TRPV1 KO mice receiving saline was similar, the values respectively being 0.5 ± 0.1 and 0.5 ± 0.2 per minute (Figs 2A,C,3A). No differences were found in the AUC between the two groups (Fig. 3C). Treatment with NGF significantly increased the frequency of voiding contractions in WT mice to 1.2 ± 0.4 per minute (P < 0.05 vs saline treatment; Figs 2B,3A). In NGF-treated KO mice, the frequency of voiding contractions was 0.5 ± 0.1 per minute (Figs 2D,3A). The amplitude of voiding contractions of WT and TRPV1 KO mice treated with saline were 27.0 ± 5.1 cmH2O and 23.4 ± 4.2 cmH2O, respectively (Fig. 3B). Repeated NGF administration both in WT and TRPV1 KO mice did not alter the amplitude of voiding contractions, the values being 23.5 ± 4.8 cmH2O and 22.8 ± 2.3 cmH2O, respectively (Fig. 3B). Prolonged NGF treatment also resulted in an increase of the AUC, both in WT mice (P < 0.05 vs saline treatment) and KO mice (Fig. 3C).
We also found significant expression of the surrogate marker of noxious sensory input c-Fos in neuronal nuclei in L6 spinal sections. The number of immunoreactive cells was very low in control WT and TRPV1 KO mice (13.4 ± 1.4 and 18.4 ± 4.0; Fig. 4). Treatment with NGF significantly increased the number of positive nuclei in WT animals (34.6 ± 0.5; P < 0.05 vs saline-treated WT mice; Fig. 3) but not in TRPV1 KO mice (23.3 ± 7.3; Fig. 4).
Expression of TrkA in the L6 segment of the spinal cord and respective DRG was assessed by immunohistochemistry. TrkA staining intensity was similar between WT and TRPV1 KO mice both in the L6 spinal cord segment (Fig. 5A,C,E) and DRG (Fig. 5B,D,F). This confirmed that expression of the high-affinity NGF receptor TrkA was identical in WT and TRPV1 KO mice.
Classically, the NGF and TRPV1 systems are seen as key players in nociception and visceral sensitization under inflammatory circumstances but have largely been regarded as parallel rather than linked pathways. However, a role of TRPV1 in bladder overactivity and noxious input associated with increased exposure to exogenous NGF was found here. In fact, TRPV1 KO mice, in contrast to WT mice, did not develop thermal hypersensitivity, significant signs of altered bladder function or spinal cord c-Fos overexpression after prolonged administration of NGF. Of note, this was not the result of altered expression of NGF high-affinity TrkA receptors, which we showed to be similarly expressed in WT and KO mice. Hence, NGF-induced bladder overactivity and increased noxious input depend on the interaction of NGF with TRPV1. This indicates that the NGF/TRPV1 interaction, shown after acute NGF administration in somatic tissues , also stands after prolonged NGF administration in a visceral model of bladder overactivity and pain. Of note, bladder changes induced by repeated NGF administration were accompanied by the development of thermal hyperalgesia, a novel fact described here.
Previous models of TRPV1 sensitization by G-protein-coupled receptor agonists, such as those for prostaglandins and bradykinin, have implied phosphatidyl-inositol-4,5-bisphosphate (PIP2) degradation and protein kinase C activation in TRPV1 regulation [12,15]. PIP2 is a molecule that maintains TRPV1 under tonic inhibition. Interestingly, PIP2 cleavage can be enhanced by NGF. Activation of TrkA receptors by NGF may lead to phospholipase C. Moreover, binding of NGF to TrkA also leads to activation of the phosphatidylinositol-3-kinase and extracellular signal-regulated kinase 1 and 2 pathways, which further results in facilitation of TRPV1 activity [16,17]. Whereas these are believed to be short-term cellular responses induced by acute administration to NGF, our findings suggest that they remain operative upon prolonged exposure to elevated NGF concentrations.
TRPV1 is unlikely to mediate all NGF responses and, similarly, NGF may not be the only pathway inducing TRPV1 sensitization [16,18,19]. In fact, the lipidic inflammatory mediators N-arachidonoyl-ethanolamine, also known as anandamide, N-arachidonoyl-dopamine, N-oleoyldopamine, eicosanoid acids and leukotrienes may also participate in TRPV1 sensitization . In addition, it should be recalled that a decrease in the inflamed tissue pH is known to reduce the heat threshold of TRPV1 from around 43 °C down to physiological temperatures . Future studies may be relevant to elucidate a possible cumulative relation between lipid mediators, protons and NGF in the process of TRPV1 activation.
Another mechanism of TRPV1 sensitization comprises down-regulation on the expression of inhibitory TRPV1 splice variants. Using the cyclophosphamide model of bladder inflammation, Charrua et al.  showed that the expression of the dominant negative splice variant TRPV1b was decreased in sensory afferents innervating the bladder. Although tempting, it cannot be concluded that NGF regulates TRPV1 alternative splicing.
There is still some debate regarding the role of TRPV1 for normal micturition. In the present study, we observed the presence of non-voiding oscillations of the bladder wall that preceded voiding bladder contractions. This is in agreement with observations made by other investigators , who also reported the presence of similar non-voiding contractions. However, it contrasts with previous results from our own group . This is probably the result of the use of a less sensitive pressure detector, which was not used in the present study. The true reasons can only be speculated but may be ascribed to compensatory responses of the bladder or the nervous system to the congenital absence of the TRPV1 receptor. It is possible that the natural presence of the receptor may dampen those non-voiding contractions by modulation of bladder sensory afferents, urothelial cells or detrusor muscle fibres . Despite the presence of these non-voiding contractions, bladder function was normal as indicated by similar frequency of voiding bladder contractions and AUC. This indicates that TRPV1 should be seen as a receptor essential for bladder dysfunction mainly related with inflammation and plays a modest role in normal bladder function [1,24].
Our findings may have profound implications in re-directing therapeutic research. There are high expectations of the use of agents interfering with NGF signalling, namely the use of tanezumab, an anti-NGF monoclonal antibody that prevents this neurotrophin from binding to its cognate receptor TrkA. A recent phase 2 study in patients with interstitial cystitis/bladder pain syndrome showed that NGF sequestration improved bladder pain at a high toll of adverse events, which included vertigo, paraesthesia and hyperesthesia . In addition, in other trials with the same drug several subjects developed bone necrosis requiring total joint replacement . This led the Food and Drug Administration to suspend the clinical trials involving tanezumab. In this scenario, TRPV1 antagonists appear as a much more attractive therapy. These drugs effectively improve bladder overactivity and noxious input associated with bladder inflammation [24,27,28]. Nevertheless, they still present some drawbacks, such as the risk of hyperthermia and increasing the extension of ischaemic tissue after coronary obliteration.
In conclusion, our results indicate that the interaction between NGF and TRPV1 is crucial for visceral overactivity and pain associated with prolonged exposure to NGF. TRPV1 therefore serves as an important bottleneck for chronic inflammatory pain, as well as visceral pain and lower urinary tract symptoms, a major reason why patients seek medical help. In this context, the development of TRPV1 antagonists assumes a clear therapeutic interest.
Financial support was given by InComb FP7 HEALTH project no 223234; Barbara Frias is supported by an FCT scholarship SFRH/BD/63225/2009 from Fundação para a Ciência e Tecnologia (Portugal).
CONFLICT OF INTEREST
Martin C. Michel is an employee of Boehringer Ingelheim. Francisco Cruz is a consultant for Astellas.