Activation of TRPV1 by capsaicin induces functional Kinin B1 receptor in rat spinal cord microglia

Background The kinin B1 receptor (B1R) is upregulated by pro-inflammatory cytokines and oxydative stress, which are enhanced by transient receptor potential vanilloid subtype 1 (TRPV1) activation. To examine the link between TRPV1 and B1R in inflammatory pain, this study aimed to determine the ability of TRPV1 to regulate microglial B1R expression in the spinal cord dorsal horn, and the underlying mechanism. Methods B1R expression (mRNA, protein and binding sites) was measured in cervical, thoracic and lumbar spinal cord in response to TRPV1 activation by systemic capsaicin (1-50 mg/kg, s.c) in rats pre-treated with TRPV1 antagonists (capsazepine or SB-366791), the antioxidant N-acetyl-L-cysteine (NAC), or vehicle. B1R function was assessed using a tail-flick test after intrathecal (i.t.) injection of a selective B1R agonist (des-Arg9-BK), and its microglial localization was investigated by confocal microscopy with the selective fluorescent B1R agonist, [Nα-bodipy]-des-Arg9-BK. The effect of i.t. capsaicin (1 μg/site) was also investigated. Results Capsaicin (10 to 50 mg/kg, s.c.) enhanced time-dependently (0-24h) B1R mRNA levels in the lumbar spinal cord; this effect was prevented by capsazepine (10 mg/kg, i.p.; 10 μg/site, i.t.) and SB-366791 (1 mg/kg, i.p.; 30 μg/site, i.t.). Increases of B1R mRNA were correlated with IL-1β mRNA levels, and they were significantly less in cervical and thoracic spinal cord. Intrathecal capsaicin (1 μg/site) also enhanced B1R mRNA in lumbar spinal cord. NAC (1 g/kg/d × 7 days) prevented B1R up-regulation, superoxide anion production and NF-kB activation induced by capsaicin (15 mg/kg). Des-Arg9-BK (9.6 nmol/site, i.t.) decreased by 25-30% the nociceptive threshold at 1 min post-injection in capsaicin-treated rats (10-50 mg/kg) while it was without effect in control rats. Des-Arg9-BK-induced thermal hyperalgesia was blocked by capsazepine, SB-366791 and by antagonists/inhibitors of B1R (SSR240612, 10 mg/kg, p.o.), glutamate NMDA receptor (DL-AP5, 10 μg/site, i.t.), substance P NK-1 receptor (RP-67580, 10 μg/site, i.t.) and nitric oxide synthase (L-NNA, 10 μg/site, i.t.). The B1R fluorescent agonist was co-localized with an immunomarker of microglia (Iba-1) in spinal cord dorsal horn of capsaicin-treated rats. Conclusion This study highlights a new mechanism for B1R induction via TRPV1 activation and establishes a link between these two pro-nociceptive receptors in inflammatory pain.


Background
Kinins are neuroactive peptides involved in pain and inflammation [1][2][3][4]. They act through the activation of two G-protein-coupled receptors (GPCR) denoted as B 1 (B 1 R) and B 2 (B 2 R) [5,6]. The B 2 R, activated by bradykinin (BK) and Lys-BK, is widely and constitutively expressed in central and peripheral tissues. The BK metabolites, des-Arg 9 -BK and Lys-des-Arg 10 -BK, are the preferential agonists of B 1 R. Whereas the B 1 R is virtually absent in healthy conditions, it is upregulated after exposure to pro-inflammatory cytokines, bacterial endotoxins and hyperglycaemia-induced oxidative stress [7][8][9][10]. The induction of B 1 R involves the transcriptional nuclear factor NF-B and MAP-kinase/P38 pathways [6,11]. We have reported that spinal injection of B 1 R agonist causes transient thermal hyperalgesia in type 1 diabetic rats due to release of sensory pro-inflammatory mediators, notably substance P (SP), prostaglandins and nitric oxide [1]. Furthermore, B 1 R antagonists reverse thermal hyperalgesia and allodynia in various models of type 1 and type 2 diabetes [4,[12][13][14][15].
The transient receptor potential vanilloid subtype 1 (TRPV1) is known as a non-selective cationic channel expressed in primary sensory C-fibers [16] and microglia [17]. Its activation increases both calcium and sodium influx [16]. TRPV1 knockout mice do not display thermal hyperalgesia [18][19][20]. TRPV1 can be sensitized through the phosphorylation of its C-terminal end by protein kinases A and/or C [21,22]. It is activated by a variety of stimuli such as heat > 43°C [16], acidification [23], BK [24], nerve growth factor [24] and oxidative stress [25]. It was recently shown that TRPV1 activation by capsaicin increases reactive oxygen species (ROS) production in mouse dorsal root ganglion (DRG) neurons [26]. TRPV1-induced ROS production is thought to involve increased cytosolic calcium influx and activation of NADPH oxidase [27]. Moreover, it has been suggested that selective TRPV1 inhibition reduces the prooxidant capacity of microglial NADPH oxidase [28].
This study was undertaken to determine whether TRPV1 activation by capsaicin could enhance expression of the pro-nociceptive B 1 R since both receptors are involved in thermal hyperalgesia. Moreover, microglial TRPV1 activation enhances pro-inflammatory cytokines and oxidative stress, both known to trigger B 1 R induction through the NF-B pathway. Thus, microglia can be considered to be a strategic target for B 1 R expression as evidenced in a diabetic model of pain neuropathy [29,30]. Our main objectives were to determine: 1-the role of oxidative stress and pro-inflammatory cytokines in capsaicin-induced B 1 R upregulation; 2-whether newly induced B 1 R is functional and could induce thermal hyperalgesia through release of spinal cord mediators; and 3-the presence of B 1 R on microglia in the spinal dorsal horn of capsaicin-treated rats by confocal microscopy.

Experimental animals and care
All research procedures and the care of the animals were in compliance with the guidelines of the Committee for Research and Ethical Issues of IASP and were approved by the Animal Care Committees of Université de Montréal and Pontificia Universidade Católica do Rio Grande do Sul. Male Sprague-Dawley rats (200-225 g; Charles River, St-Constant, Qc, Canada and CEMIB, UNICAMP, Brasil) were housed two per cage, under controlled conditions of temperature (23°C) and humidity (50%), on a 12 h light-dark cycle (until surgery) and allowed free access to normal chow diet (Charles River Rodent) and tap water.

Intrathecal implantation of catheter and capsaicin treatment
Four days after arrival, rats were anaesthetized with isoflurane and chronically implanted with an indwelling intrathecal (i.t.) polyethylene catheter (PE-10; Intramedic, Clay Adams, NJ, USA) at the vertebral lumbar level (L3 to L6) through an incision made in the dura at the atlanto-occipital junction [31]. The canula was secured to the skull through two small parallel segments of PE-60 glued with cyano-acrylate. Rats with apparent abnormal behaviour or motor deficits were euthanized with CO 2 inhalation. Rats were housed permanently in the testing laboratory under continuous light to prevent the release of endogenous opioids which could alter nociceptive threshold [32]. One week after surgery, six groups of 7-8 rats received a single dose of capsaicin (1, 5, 10, 15, 25 and 50 mg/kg), injected subcutaneously at the lumbar back skin level under mild anaesthesia with pentobarbital (45 mg/kg, i.p.). The highest dose of 50 mg/kg was injected in two doses of 25 mg/kg at 12 h intervals. Rats were used 24 h after capsaicin injection, except for rats used for time-course experiments (treated for: 0, 8, 24 or 48 h). To evaluate directly the capacity of spinally expressed TRPV1 to induce B 1 R expression, one group of rats received 1 μg of capsaicin intrathecally at the lumbar region of the spinal cord. Control rats received vehicle only (10% ethanol, 10% Tween-80 and 80% saline 0.9%). Responses to des-Arg 9 -BK were measured using a tail-flick test prior to and 24 h after capsaicin injection.

Tail-flick test protocol
Experiments were always started in the morning at around 10:00 H AM. Awake rats were placed in a plastic restraining box. The nociceptive threshold was taken as the reaction-time to remove the tail from a source of noxious heat stimulus using either a tungsten lamp or a hot water bath (50 ± 1°C) [33]. The average tail-flick reaction times of control rats were 9.4 ± 0.2 s and 5.8 ± 0.3 s, when assessed with the tungsten lamp and the hot water bath as heat stimuli, respectively. A 25 s cut-off time was used to prevent tissue damage [31]. All rats were tested for a maximum of four consecutive days. During the first two days, rats received 10 μl of artificial cerebrospinal fluid (aCSF) as training experiments and to ensure that the intrathecal catheter was patent. On the subsequent days, each testing trial lasted 45 min and consisted of 9 measurements of tail-flick latency, spaced by 5 min intervals. The initial three measurements were used to determine baseline latency [31]. One minute prior to the 4 th reading (t = 15 min), the vehicle (aCSF) or an inhibitor was i.t. injected. Twelve min later, the tested agonist (des-Arg 9 -BK) or substance P (SP) (as positive control) was administered through the same route (t = 26 min), and its effect on the tail-flick latency was measured 1 min later (t = 27 min). Three subsequent readings were made to assess the effect of the agonist on the nociceptive threshold. The study design for time and agonist doses (des-Arg 9 -BK, 9.6 nmol; SP, 6.6 nmol) was based on previous studies showing that SP and des-Arg 9 -BK evoke a transient 5-10 min thermal hyperalgesia that peaks 1 min post-injection [1,31,34]. Drugs and tested agonists were intrathecally injected using a 50 μl Hamilton syringe with a total volume of 10 μl. The catheter, with a void volume of 10 μl, was immediately flushed after drug injection by the administration of 15 μl aCSF.

Antioxidant treatment
Rats without intrathecal catheter received every day for a period of 1-week the potent antioxidant N-acetyl-Lcysteine (NAC) (1 g/kg/d) or the vehicle (sterile water) by gavage [8]. On the 6 th day of treatment with NAC, rats received a single injection of capsaicin (15 mg/kg, s. c.) or its vehicle. Rats were sacrificed 24 h later under inhalation with CO 2 .

Tissue preparation for autoradiography and microscopy
Twenty-four hours after capsaicin treatment, rats were anaesthetised with CO 2 inhalation and then decapitated. Lower-lumbar (L3-L6) spinal cord was removed and frozen in 2-methylbutane (cooled at -55 ± 5°C with liquid nitrogen) and stored at -80°C. Spinal cords were mounted in a gelatin block and serially cut into 20-μm thick coronal sections with a cryostat. The sections were thaw-mounted on 0.2% gelatin-0.033% chromium potassium sulfate-coated slides and kept at -80°C for 1 month to allow sections adhesion to the coverslip glasses.

Quantitative autoradiography
The density of kinin B 1 R binding sites was measured with the radioligand [ 125 I]-HPP-desArg 10 -Hoe140 (3-(4 hydroxyphenyl) propionyl-desArg 9 -D-Arg 0 [Hyp 3 , Thi 5 , D-Tic 7 , Oic 8 ]bradykinin) used at a concentration of 200 pM (specific activity: 2000 cpm/fmol) as described previously [3,8,15,29]. Non-specific binding was determined for each capsaicin doses in the presence (1 μM) of the selective B 1 R antagonist R-715 (AcLys[D-βNal 7 , Ile 8 ]des-Arg 9 -BK) [45]. Scientific Imaging Films BIOMAX ™ MR ® (Amersham Pharmacia Biotech Canada) were juxtaposed onto the slides in the presence of [ 125 I]-microscales and exposed at room temperature for 7 days. Autoradiograms were quantified by densitometry using an MCID™ image analysis system (Imaging Research, St. Catharines, ON, Canada). Briefly, twenty measurements were made throughout the spinal cord dorsal horn in 3 to 8 rats per group. A standard curve from [ 125 I]-microscales was used to convert density levels into fentomoles per milligram of protein [46]. Specific binding was determined by subtracting values of nonspecific binding from that of total binding.

Measurement of superoxide anion
In situ levels of O 2 in the spinal cord were evaluated with the oxidative fluorescent dye dihydroethidium (DHE) as described earlier [47]. Cells are permeable to DHE which specifically react with O 2 to produce ethidium bromide (EtBr) which is trapped by intercalation with DNA. EtBr is excited at 518 nm with an emission spectrum of 605 nm. Twenty-four hours post-capsaicin, rats were anaesthetized with CO 2 inhalation and then decapitated. The lumbar spinal cord (L3-L6) was isolated and serially cut into 20-μm thick sections and placed on glass slides. On the day of experiment, slides were exposed for 30 min in a light-protected humidified chamber at 37°C to DHE (2 μM) for superoxide anion labeling and to TO-PRO-3 1:5000 (Molecular Probes, Eugene, OR, USA) for DNA staining. Sections were then washed 3 times in sterile phosphate buffered saline (PBS, pH7.4), coverslipped and observed using a confocal microscope (Leica Microsystem Co., Germany). Slides from the four groups (control, control + NAC, capsaicin-treated, capsaicin-treated + NAC) were processed and imaged in parallel. Laser settings were identical for acquisition of images from all sections. Computer based analysis was performed with Image J software and calculated using the following equation: I=∑ I/(A/N), where I is the fluorescence intensity, ∑ I is the summation of all nuclei intensity, A is the total area of the nuclei, and N is the number of nuclei used [47]. Data are expressed as an average of total nuclei fluorescence quantified on at least 4 nuclei/section on 4 sections/rat from 4 different rats. Positive pixels regarding the co-localisation of B 1 R and Iba-1 were determined following the subtraction of mean pixel background level intensity from total pixels.

SYBR green-based quantitative real-time PCR
Twenty-four hours after capsaicin injection, rats were anaesthetized with CO 2 inhalation and then decapitated. Cervical (C2-C4), thoracic (T3-T5) and lumbar spinal cord (L3-L6) segments were isolated and approximately 10 mg of tissue were put in RNAlater stabilization reagent (QIAGEN, Valencia, CA, USA). The protocol for mRNA extraction, cDNA generation, SYBR greenbased quantitative RT-PCR and quantification has been described elsewhere [29]. The PCR conditions were as follows: 95°C for 15 min, followed by 46 cycles at 94°C for 15 s, 60°C for 30 s and 72°C for 30 s. Real-time PCR primer pairs were designed using Vector NTI and are presented in Table 1.

Western blot
The extensive procedure has been described elsewhere [48]. Briefly, 24 h after capsaicin treatment, rats were anaesthetized with CO 2 inhalation and then decapitated. The lumbar spinal cord (L3-L6) was isolated and approximately 50 mg of tissue were put in PBS containing a cocktail of proteases inhibitors (Sigma-Aldrich, Canada). Antibodies against B 1 R, p65NF-kB and dynein (internal control) are presented in Table 2.

Drugs and solutions
Des-Arg 9 -BK and SP were purchased from Bachem Bioscience Inc. were dissolved in aCSF while SB-366791 and capsazepine were dissolved in sterile saline for i.p. injections. SSR240612 was dissolved in dimethyl sulfoxide (DMSO, 0.5% v/v), ethanol (5% v/v) and Tween-80 (5% v/v). The solution was completed in distilled water. The drug was administered orally by gavage in a volume of 1 ml per 100 g of body weight [39]. Capsaicin was solubilized in a mixture of 10% ethanol, 10% Tween-80 and 80% saline 0.9%.

Statistical analysis
Data were expressed as means ± S.E.M. of values obtained from n rats. Only one treatment was given to a rat. For the tail-flick test, data were calculated as a percentage of the maximum possible effect (% MPE) according to the following formula: % MPE = 100 × (drug latency minus baseline latency/cut-off time minus baseline latency) [1]. Statistical significance was determined with Student's t-test for paired samples or with one-way analysis of variance (ANOVA) followed by the Bonferroni post-test for multiple comparisons. Data for DHE labelling intensity were analysed with the nonparametric Kruskal-Wallis post-test. Probability (P) values less than 0.05 were considered to be statistically significant.

Increased B 1 R expression by capsaicin-induced TRPV1 activation
A 24-h-capsaicin treatment (10-50 mg/kg, s.c.) increased significantly B 1 R mRNA levels in rat lumbar spinal cord when compared with vehicle-treated rats. A dose of 15 mg/kg caused the highest increase of gene expression (500-fold) at 24 h and the effect was almost gone at 48 h ( Figure 1). In comparison, a dose of 50-mg/kg capsaicin caused a smaller enhancing effect on B 1 R mRNA, which peaked at 8 h and declined thereafter ( Figure 1B). The increasing effect of 15 mg/kg capsaicin on B 1 R mRNA was significantly less at the thoracic (T4) and cervical (C2) spinal cord levels ( Figure 2). Intrathecal capsaicin (1 μg/site) reproduced to some extent the effects of systemic capsaicin, as it significantly increased B 1 R mRNA in the lumbar spinal cord in comparison with vehicle-treated rats 24 h post-treatment ( Figure 3).
Contrarily to B 1 R mRNA levels, the density of B 1 R binding sites was dose-dependently increased at 15 and 50 mg/kg capsaicin ( Figure 5). As depicted on autoradiograms, B 1 R binding sites were displayed throughout the grey matter of spinal cord in capsaicin-treated rats and poorly in white matter. Such a distribution is in accordance with B 1 R expression on spinal cord projecting C-fibres, astrocytes and microglia [29].
Effect of des-Arg 9 -BK on tail-flick latency in capsaicintreated rats The functionality of the B 1 R which is up-regulated by capsaicin in the spinal cord was assessed on thermal nociception. Intrathecal injection of des-Arg 9 -BK (9.6 nmol/site)  failed to alter nociceptive threshold in control (vehicle) rats or in rats treated with capsaicin at doses of 1 and 5 mg/kg, 24 h earlier ( Figure 6A). However, the B 1 R agonist dose-dependently decreased the tail-flick latency in rats treated with doses from 10 to 50 mg/kg capsaicin. The hyperalgesic response to the B 1 R agonist peaked l min post-injection (-30%) and lasted about 5-10 min in rats pre-treated with 15 mg/kg capsaicin (Figure 7). The

Housing conditions
Most of the current data were obtained in rats housed under constant light with the intent to prevent the release of endogenous opioid peptides [32] which could interfere with the hyperalgesic response to substance P in the rat tail-flick test [31,34]. To confirm that the interruption of the circadian cycle by constant light did not corrupt our findings, we run a series of experiments   Figure 1A) were not significantly different from those measured in rats under the 12 h-12 h light dark-cycle ( Figure 1B). Thus, these data confirm that constant light exposure does not interfere with our findings.

Microglial localisation of B 1 R in the spinal dorsal horn of capsaicin-treated rats
As shown in Figure 10, immunofluorescence to the specific marker of microglia Iba-1 (red, A) and fluorescent labeling with the specific B 1 R agonist [N α -bodipy]-des-Arg 9 -BK (green, B) were co-localized (yellow, C) in spinal cord dorsal horn of capsaicin-treated rats (15 mg/ kg, s.c., 24 h earlier). Note that 87 ± 4% of B 1 R-positive pixels co-localized with Iba-1-positive pixels, suggesting that B 1 R induced by capsaicin is mostly expressed by lumbar spinal cord microglia.

Increased superoxide anion production and NF-B activation in the spinal cord of capsaicin-treated rats
This series of experiments aimed to determine the contribution of oxidative stress to the induction of B 1 R following TRPV1 activation. DHE staining (red staining, Figure 11C) was significantly increased in nuclei (labelled with TO-PRO-3 in blue) in the spinal dorsal horn of rats treated with capsaicin (15 mg/kg, s.c., 24 h earlier) when compared with control spinal cord ( Figure  11A). This increase was abolished by a 1-week pre-treatment with NAC (1 g/kg/d) (from 1.46 ± 0.21 to 0.74 ± 0.12 a.u.) ( Figure 11D). In contrast, NAC treatment had no significant effect on DHE staining in control spinal cord ( Figure 11B). This is in relation with the increased protein expression of B 1 R in the spinal cord of capsaicin-treated rats, which was also prevented by pre-treatment with NAC ( Figure 12). These data thus suggest that oxidative stress is involved in the induction of B 1 R by TRPV1 stimulation. To determine the contribution of NF-B in this process, expression of p65NF-B was determined. Capsaicin-treatment caused a non-significant increase of p65NF-B when compared with control spinal cords. However, NAC treatment (1 g/kg/d × 7 days) reduced significantly p65NF-B expression in the spinal cord of capsaicin-treated rats. NAC had no effect in control spinal cord p65NF-B expression ( Figure 12).

Changes of IL-1b and TNF-a mRNA levels in the spinal cord of capsaicin-treated rats
To further determine the mechanism of B 1 R induction by capsaicin, spinal cord mRNA levels of pro-inflammatory cytokines (IL-1β and TNF-α) known to induce B 1 R were assessed using real-time PCR ( Figure 13). These data show a significant increase in IL-1β mRNA expression (40-fold) in the spinal cord of capsaicin-treated rats (15 mg/kg, s.c.) when compared with control. This induction was prevented by the selective TRPV1 antagonist capsazepine (10 mg/kg, i.p.) administered 1 h prior to capsaicin. In contrast, TNF-α mRNA levels were unaffected by the three doses of capsaicin. Spinal cord B 1 R and IL-1β mRNA levels following capsaicin challenge were highly correlated (R 2 = 0.996) while mRNA levels of B 1 R and TNF-α (R 2 = 0.562) were not correlated.

Discussion
The present study provides the first pharmacological evidence that the B 1 R (mRNA and protein levels) can be induced and up-regulated in rat spinal cord after systemic or spinal activation of TRPV1 by capsaicin. The B 1 R induction mechanism involves oxidative stress, proinflammatory cytokines and the NF-B pathway.
Activation of TRPV1 on primary sensory afferents and microglial cells can enhance superoxide anion production [26,28], which is known to induce B 1 R expression in various tissues [4,8,15]. Hence, the involvement of oxidative stress in the induction of B 1 R by TRPV1 is consistent with these latter studies, as shown with Nacetyl-L-cysteine treatment that prevented the increased expression of B 1 R and superoxide anion levels in the spinal cord of capsaicin-treated rats. Functionality of the B 1 R was demonstrated in thermonociception. The hyperalgesic response to intrathecally injected B 1 R agonist was ascribed to the intraspinal release of NO and activation of NK-1R and NMDA-R, as previously reported following spinal activation of B 1 R in streptozotocin-diabetic rats [1]. The presence of B 1 R in spinal cord microglia is consistent with the emerging role of microglial B 1 R in pain neuropathy [29,30,49].

TRPV1 activation and inhibition
In this study, TRPV1 was activated with increasing doses of the lipophilic molecule capsaicin, the active pungent ingredient of hot chilli peppers [50]. Capsaicin's actions are classified into TRPV1-mediated actions (including some side effects) and TRPV1-independent effects. TRPV1-mediated side effects include desensitisation (depletion of SP and CGRP) of C and Aδ neurons following sub-chronic capsaicin treatment, and neuronal apoptosis (calcium neurotoxicity) after chronic treatment with high capsaicin doses [50]. As reviewed by the latter authors, capsaicin TRPV1-independent effects include: alteration of membrane fluidity [51], inhibition of platelet aggregation [52], and toxicity to non-TRPV1 neurons [53]. Hence, to ascertain that the enhancing effect of capsaicin on B 1 R expression is truly mediated by TRPV1 activation, two classes of TRPV1 antagonists (capsazepine and SB-366791) were used [16]. Although high doses of capsazepine can block neuronal voltagegated calcium channels [54], it has also been shown that capsazepine is 10-100 times more potent for TRPV1 than for its off-target [54]. Secondly, we selected SB-366791 which is known as a more selective and more potent TRPV1 antagonist than capsazepine [42].
Recently, it has been broadly employed as a selective TRPV1 antagonist in pain research [55]. In addition to blocking B 1 R up-regulation induced by capsaicin, capsazepine and SB-366791 reversed selectively des-Arg 9 -BKinduced hyperalgesia without affecting SP-induced hyperalgesia. As additional evidence of specificity, systemic and intrathecal treatments with TRPV1 antagonists (capsazepine and SB-366791) had no direct effect on baseline latency in the tail-flick test. Collectively, these findings strongly suggest that the induction of B 1 R by capsaicin is attributable to TRPV1 stimulation.

Localisation of B 1 R and site of action for capsaicin
Using a fluorescent B 1 R ligand, B 1 R was mostly found (87% of positive cells) in microglial cells in the lumbar spinal cord dorsal horn of capsaicin-treated rats. Such a cellular localisation on migratory cells can explain the widespread distribution of B 1 R binding sites in all spinal cord laminae. The significant increase of B 1 R mRNA in the spinal cord of capsaicin-treated rats also supports a spinal site rather than a peripheral site (DRG) for B 1 R synthesis. The induction of B 1 R was largely restricted to lumbar spinal cord segments innervated by peripheral nociceptors in the stimulated region, suggesting a contribution of peripheral TRPV1 to the induction of B 1 R in the spinal cord. The enhanced expression of B 1 R on rostral segments (T4 and C2) of the spinal cord could be due to the expression of B 1 R on migratory microglial cells after stimulation of sensory fibers by capsaicin.
Our data also suggest that induction of B 1 R by capsaicin is mediated by a central mechanism since TRPV1 antagonists administered intrathecally prevented B 1 R mRNA expression and B 1 R-stimulated thermal hyperalgesia. Capsaicin and capsazepine are lipophilic molecules and SB-366791 is likely to pass the blood-brain barrier as well [56]. Therefore, systemically administered capsaicin is thought to enter the spinal cord to stimulate TRPV1 located on central sensory terminals, astrocytes [56] and microglia [56]. Supporting the latter assumption, we showed that 1 μg of capsaicin directly injected into lumbar spinal cord caused a significant increase in B 1 R mRNA, showing that centrally expressed TRPV1 cells can trigger B 1 R expression.

B 1 R mRNA expression patterns
Contrarily to B 1 R mRNA expression, B 1 R specific binding sites were shown to be maximally increased by 50 mg/kg capsaicin. This is congruent with the faster kinetics of B 1 R mRNA expression after a dose of 50 mg/ kg capsaicin in comparison with 15 mg/kg. As B 1 R is hardly internalised or desensitised [6], its presence on the cellular membrane can easily outlast the transient increase of mRNA in response to acute stimulation of TRPV1. Thus, B 1 R mRNA changes do not reflect the expression of B 1 R on the cellular membrane. The actual bell-shaped pattern of B 1 R mRNA expression is not unique and has been observed in other models of acute inflammation induced by exposure to increasing concentrations of tobacco-smoke condensate [48,57].
Mechanism underlying B 1 R induction TRPV1-induced oxidative stress is suggested as the primary mechanism by which capsaicin induces B 1 R. This is supported by the suppression of capsaicin-induced increase B 1 R protein expression and NF-kB activation by the 1-week treatment with the antioxidant N-acetyl-L-cysteine. This is in line with the findings of Ma et al. [26], which suggest that TRPV1 activation by capsaicin increases reactive oxygen species production in mouse dorsal root ganglion neurons. This can occur through increased cytosolic calcium influx and activation of NADPH oxidase [27]. N-acetyl-L-cysteine also prevented the increased superoxide anion production in the spinal cord of capsaicin-treated rats. It is noteworthy that Nacetyl-L-cysteine or a prolonged treatment with another antioxidant (alpha-lipoic acid) has been shown to prevent rat spinal cord B 1 R induction in models of diabetes [8,15] and hypertension [58]. In a model of rat knee joint-induced arthritis, peripheral TRPV1 and centrally increased oxidative stress can enhance pro-inflammatory cytokines production [59], suggesting that these inflammatory molecules can also contribute to the effect of capsaicin on B 1 R expression. Indeed, stimulation of TRPV1 causes the release of various pro-inflammatory cytokines, including interleukin-6 (IL-6) and IL-8 from human bronchial epithelial cells [60], IL-1ß and transforming growth factor-ß 2 from human ORS keratinocytes [61], IL-2, IL-4 and interferon-gamma from cultured murine Peyer's patch cells [62], or IL-6 release and NF-B activation in dorsal horn TRPV1-expressing microglia [63]. Those pro-inflammatory cytokines could activate NF-B translocation to the nucleus, thereby increasing B 1 R expression [1,64,65]. One can suggest that cytokines may represent key mediators in the induction of B 1 R following TRPV1 activation. As indirect evidence, we found that IL-1β mRNA levels are enhanced and highly correlated with B 1 R mRNA levels following TRPV1 stimulation. In contrast, TNF-α is unlikely to be involved in B 1 R expression as its mRNA level was unaffected by capsaicin administration. This is in agreement with data showing no correlation between B 1 R induction and TNF-α expression in tobacco smokeinduced lung inflammation [48]. B 1 R activation led to thermal hyperalgesia TRPV1 activation by capsaicin induced functional B 1 R in the rat spinal cord. This is highlighted by the hyperalgesia induced by intrathecal administration of B 1 R selective agonist (des-Arg 9 -BK) in capsaicin-treated rats. This response was inhibited by SSR240612, a highly selective and orally active B 1 R antagonist [38]. In contrast, des-Arg 9 -BK had no effect on the nociceptive threshold in control rats which is congruent with its weak constitutive expression. The spinal cord activation of tachykinin NK-1 receptor with SP causes a hyperalgesic response in the tail-flick test which is associated with glutamate release and NO production. Indeed, blockade of NOS or NMDA receptor prevents thermal hyperalgesia induced by endogenous release of SP following noxious cutaneous stimulation or by intrathecal administration of SP [66][67][68]. The hyperalgesic response induced by SP has the same time-course and amplitude as that induced by des-Arg 9 -BK. Also, our pharmacological analysis shows that the hyperalgesic response to B 1 R agonist is likely mediated by endogenous release of NO and activation of NK-1R and NMDA-R. These spinal pro-nociceptive mediators appear to be tonically active in capsaicin-treated rats as their inhibition increased the nociceptive threshold. Hypoalgesia has also been reported with inhibition of NOS with L-NNA in streptozotocin-diabetic rats [1] or with NMDA receptor blockade [69]. Since the hyperalgesic response to SP in capsaicin-treated rats was not significantly altered, the hyperalgesic response to des-Arg 9 -BK is unlikely to be due to an upregulation of NK-1 receptors or increased SP response in the spinal cord of capsaicin-treated rats. The TRPV1 antagonists capsazepine and SB-366791 prevented the hyperalgesia induced by des-Arg 9 -BK but not that evoked by SP, suggesting that this inhibition is highly specific.

Conclusion
This study provides the first evidence that B 1 R can be induced and up-regulated in microglia of spinal cord dorsal horn following activation of TRPV1 by capsaicin. This up-regulation is correlated with increased expression of IL-1β and involves ROS generation and the redox-regulated NF-B pathway. The newly synthesized B 1 R is functional as its activation with an agonist caused thermal hyperalgesia which we attribute to the intraspinal release of NO and activation of NK-1R and NMDA-R. These findings support the possibility that microglial B 1 R contributes to the effect of TRPV1 in inflammatory pain.  Changes in IL-1b and TNF-a mRNA levels in lumbar spinal cord of capsaicin-treated rats (0-50 mg/kg, s.c. 24 h earlier). Also shown is the effect of the TRPV1 antagonist capsazepine (10 mg/kg, i.p.) administered 1 h prior to capsaicin. Data represent the means ± S.E.M. of 5-7 rats in each group. Comparison with vehicle (0 mg/kg) (*) is indicated by * P < 0.05.