Role of transient receptor potential vanilloid 4 in rat joint inflammation
University of São Paulo and Federal University of São Paulo, São Paulo, Brazil, and INSERM, U1043, Université Paul Sabatier Toulouse, Centre de Physiopathologie de Toulouse Purpan, and CNRS, U5282, Toulouse, France
To determine whether activation of transient receptor potential vanilloid 4 (TRPV-4) induces inflammation in the rat temporomandibular joint (TMJ), and to assess the effects of TRPV-4 agonists and proinflammatory mediators, such as a protease-activated receptor 2 (PAR-2) agonist, on TRPV-4 responses.
Four hours after intraarticular injection of carrageenan into the rat joints, expression of TRPV-4 and PAR-2 in trigeminal ganglion (TG) neurons and in the TMJs were evaluated by real-time reverse transcription–polymerase chain reaction and immunofluorescence, followed by confocal microscopy. The functionality of TRPV-4 and its sensitization by a PAR-2–activating peptide (PAR-2–AP) were analyzed by measuring the intracellular Ca2+ concentration in TMJ fibroblast-like synovial cells or TG neurons. Plasma extravasation, myeloperoxidase activity, and the head-withdrawal threshold (index of mechanical allodynia) were evaluated after intraarticular injection of selective TRPV-4 agonists, either injected alone or coinjected with PAR-2–AP.
In the rat TMJs, TRPV-4 and PAR-2 expression levels were up-regulated after the induction of inflammation. Two TRPV-4 agonists specifically activated calcium influx in TMJ fibroblast-like synovial cells or TG neurons. In vivo, the agonists triggered dose-dependent increases in plasma extravasation, myeloperoxidase activity, and mechanical allodynia. In synovial cells or TG neurons, pretreatment with PAR-2–AP potentiated a TRPV-4 agonist–induced increase in [Ca2+]i. In addition, TRPV-4 agonist–induced inflammation was potentiated by PAR-2–AP in vivo.
In this rat model, TRPV-4 is expressed and functional in TG neurons and synovial cells, and activation of TRPV-4 in vivo causes inflammation in the TMJ. Proinflammatory mediators, such as PAR-2 agonists, sensitize the activity of TRPV-4. These results identify TRPV-4 as an important signal of inflammation in the joint.
Transient receptor potential vanilloid 4 (TRPV-4), a cation channel of the TRP family, is a membrane environmental sensor responsive to thermal, osmotic, chemical, and mechanical stimuli. TRPV-4 can be directly gated by 5′,6′-epoxyeicosatrienoic acids derived from anandamide or arachidonic acid metabolism (1–3), by semisynthetic compounds such as 4α-phorbol 12,13-didecanoate (4α-PDD) (4), or by synthetic compounds such as GSK1016790A (5). It is widely expressed in vascular, musculoskeletal, and sensory systems (5–7). Mutations in the Trpv4 gene, leading to gain-of-function, have been associated with several human musculoskeletal and sensory disorders (8, 9). In contrast, loss-of-function in Trpv4-knockout mice or inherited single-point mutations in this gene in humans, leading to unresponsiveness to hypo-osmotic stimuli, will induce severe arthropathies, suggesting that this channel plays a chondroprotective role (10). Overall, musculoskeletal homeostasis seems to be highly dependent on a tight regulation of TRPV-4, since up- or down-regulation of its activity can be deleterious.
The polymodal sensory capabilities of this receptor and its expression in sensory systems suggest that TRPV-4 plays a role in molecular mechanotransduction (5, 11). Indeed, TRPV-4–deficient mice showed an increased somatic mechanical nociceptive threshold (11). TRPV-4 expressed within sensory neurons can be sensitized by proinflammatory mediators, such as prostaglandin E2, histamine, and/or serotonin, leading to increased nociception to hypotonic or mild hypertonic stimuli or even mechanical stimuli (12–15). In addition to the mediators described above, activation of the proteinase-activated receptor 2 (PAR-2) is a strong signal that sensitizes TRPV-4 (16). Interestingly, development of experimental arthritis is impaired in F2rl1 (PAR-2)–knockout mice (17); thus, PAR-2 may be proposed as a putative therapeutic target for arthritis (18).
We have recently shown that PAR-2 activation within the rat temporomandibular joint (TMJ) induced inflammation via a neurogenic mechanism (19). The pronociceptive effects of PAR-2 activation are tightly dependent on TRPV-4–mediated Ca2+ influx (16, 20), which indicates that there is a potential interaction between these receptors. Furthermore, TRPV-4 activation in the mouse paw induced edema and granulocyte recruitment by stimulating the release of proinflammatory neuropeptides, such as calcitonin gene-related peptide and substance P, from sensory nerve fibers (21).
Based on these findings implicating a role for TRPV-4 in musculoskeletal disorders and showing a possible linkage between PAR-2 and TRPV-4, we hypothesized that TRPV-4 plays a major role in joint inflammation. Thus, in the present study, we assessed the pattern of expression of TRPV-4 in the TMJ and in trigeminal ganglion (TG) neurons of rats, and studied its regulation during carrageenan-induced inflammation. The function of TRPV-4 was assessed in vitro by measuring the changes in Ca2+ influx in primary cultures of TG neurons and TMJ fibroblast-like synovial cells. In vivo, parameters of inflammation and nociception were assessed. Similarly, sensitization of TRPV-4 by PAR-2 activation was considered in vitro and in vivo by assessing the coexpression and function of these 2 receptors.
MATERIALS AND METHODS
Male Wistar rats (weighing 200–250 gm) were obtained from Janvier (Le Genest Saint Isle, France) or the Institute of Biomedical Sciences of the University of São Paulo (Brazil). Institutional Animal Care Committees approved the experimental protocols. All the injections were performed in animals anesthetized with 5% isoflurane or ketamine and xylazine (80 mg/kg and 20 mg/kg, respectively) intraperitoneally.
The PAR-2 agonist SLIGRL-NH2, a PAR-2–activating peptide (PAR-2–AP), and control reverse peptide LRGILS-NH2 (PAR-2–RP) were from the Peptide Core Facility of the University of Calgary (Calgary, Alberta, Canada), and stock solutions (100 mM) were prepared in Hanks' balanced salt solution (HBSS; Gibco). The TRPV-4 agonist 4α-PDD was supplied by Calbiochem-VWR, and stock solutions (7.5 mM) were prepared in DMSO. The TRPV-4 agonist GSK1016790A (Sigma) was prepared in DMSO–15 mM stock solution. Fluo-3 AM (Molecular Probes/Invitrogen) was prepared in DMSO. Probenecid (500 mM in 1M NaOH), pluronic F-127 (20% in DMSO), papain, Dispase II, type IA collagenase, trypsin IX-S, L-cysteine, poly-L-ornithine, laminin, 5-fluoro-2′-deoxyuridine, uridine, fetal bovine serum (FBS), EGTA, carrageenan, and DMSO were purchased from Sigma-Aldrich.
Induction of TMJ inflammation with carrageenan.
Acute inflammation in the TMJ was induced by an intraarticular injection of carrageenan (10 μl of a 5% carrageenan solution in sterile saline) into the supradiscal space of the left TMJ, using a microsyringe (Hamilton) coupled to a 30-gauge gingival needle (Terumo). In addition, the same volume of vehicle was injected into the left TMJ of animals belonging to the control group. Rats were anesthetized and then killed by decapitation 4 hours after the induction of inflammation.
Immunofluorescence and confocal microscopy.
TMJs and TG neurons were harvested from the rats, fixed in 4% formaldehyde, cryoprotected, cut into 20-μm sections in a cryostat, and mounted on a Superfrost slide (Thermo Fisher Scientific), as previously described (19). Slides were washed in phosphate buffered saline (PBS), 0.5% Triton X-100, and 1% bovine serum albumin (BSA) solution (Sigma) and incubated overnight at 4°C with the primary antibodies (1:100 anti–PAR-2, sc-8207 or sc-5597; Santa Cruz Biotechnology) along with anti–TRPV-4 and/or anti-CD68 (1:200, ab50578 and ab31630; both from Abcam). To test the specificity of the anti–PAR-2 and anti–TRPV-4 antibodies, both were previously tested in a study using dorsal root ganglia slices of corresponding gene-deficient mice (Trpv4- or F2rl1 [PAR-2]–knockout mice), and no staining was observed (results not shown).
After washing, slides were incubated with the appropriate secondary antibody conjugated with Alexa Fluor 488, Alexa Fluor 555, or Alexa Fluor 680, washed, and mounted with ProGold containing DAPI (Molecular Probes). Cells were then washed with PBS and fixed in formaldehyde. After washing, cells were permeabilized for 5 minutes in PBS plus 1% BSA and 0.05% Triton X-100, and incubated with the antibodies described above.
Images were acquired using Zeiss LSM-510 or LSM-710 confocal microscopes (Carl Zeiss MicroImaging) with 20×, 40×, or 63× objectives in the inverted configuration. The mean diameter of the neurons and the quantification of fluorescence were determined using Metamorph software (Molecular Devices).
Analysis of TRPV-4 and PAR-2 gene expression by real-time reverse transcription–polymerase chain reaction (RT-PCR).
Total RNA was extracted from the TMJ with the RNeasy Mini kit (Qiagen), and 150 ng of the total RNA was then reverse-transcribed with random hexamer oligonucleotides and SuperScript III, according to the manufacturer's instructions (Invitrogen). Total RNA was extracted from TG neurons using TRIzol reagent (Invitrogen). RNA (3 μg) was reverse-transcribed with Moloney murine leukemia virus RT, using Oligo(dT)20 for priming (Invitrogen). Amplification was performed with a LightCycler 480 using a SYBR Green I Master kit (Roche Applied Science) and the following primers: for Trpv4, forward ACCACGGTGGACTACCTGAG and reverse AGCCATCGACGAAGAGAGAA (accession number NM_023970.1); for F2rl1 (PAR-2), forward TGGGAGGTATCACCCTTCTG and reverse AAAAGCCTGGTTCAACTGGA (accession number NM_053897.2); and for Hprt1, forward AAGCTTGCTGGTGAAAAGGA and reverse TGATTCAAATCCCTGAAGTGC (accession number NM_012583.2).
Each amplification reaction was performed in triplicate, with negative controls consisting of complementary DNA substituted by the RT-PCR product of reactions carried out without the RT enzyme. Melting curve analysis, conducted at temperatures of 76–84°C, was performed at the end of each run as a quality-control step. Relative expression of the target gene was normalized to expression of the Hprt gene, using the ΔΔCt method (22).
Dispersion and culture of rat TG neurons.
TG neurons harvested from rats were digested in 27 μg/ml papain for 10 minutes at 37°C. Samples were washed in prewarmed Leibovitz's L-15 medium with 2 mM glutamine and 10% FBS, and further digested in HBSS containing 1 mg/ml of type IA collagenase and 4 mg/ml of Dispase II for 10 minutes at 37°C. Cells were centrifuged and pellets were suspended in Dulbecco's modified Eagle's medium (DMEM) containing 10% FBS, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 2 mM glutamine, 10 μM 5-fluoro-2′-deoxyuridine, and 10 μM uridine. For Ca2+ influx assays, cells were plated at the desired confluence on 8-well Lab-Tek II plates (Nunc). Neurons were cultured for 2–3 days at 37°C in an atmosphere of 5% CO2.
Dispersion and culture of rat fibroblast-like synovial cells.
The synovium surrounding the periphery of the TMJ disc was carefully dissected on a stereomicroscope. The tissue was digested for 30 minutes at 37°C with type IX-S trypsin (1 mg/ml) added to DMEM. The tissue was then washed with DMEM alone and digested for 2 hours at 37°C in type IA collagenase (1 mg/ml) in DMEM supplemented with 10% heat-inactivated FBS. The digested tissue was then passed through a 70-μm nylon cell strainer and the cells were pelleted. The cells were then suspended in DMEM supplemented with 10% FBS, containing 25 mM HEPES, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 0.25 μg/ml amphotericin B, and 2 mM glutamine, and 2.5 × 105 cells were plated on a 75-cm2 culture flask. Cells were cultured for 5–7 days until 90% confluence was reached, and were passaged using the proteinase-free reagent Versene. For measurement of [Ca2+]i concentrations, serum-starved cells were plated in a 96-well plate (1 × 104 cells per well) and cultured for 24 hours. For immunofluorescence analysis, the same number of cells was plated in 8-well Lab-Tek II plates and cultured for 24 hours.
Cell flow cytometry.
Synoviocytes were labeled with an antibody against the macrophage marker CD68 by the indirect intracellular technique, according to the manufacturer's instructions (Abcam). Labeling was revealed with the secondary antibody anti-mouse IgG conjugated to fluorescein isothiocyanate (Jackson ImmunoResearch). The system was calibrated with nonlabeled cells, and control samples consisted of reactions performed only with the secondary antibody. The cells were resuspended in 300 μl of 0.1% formaldehyde in PBS, and results in 5,000 events per sample were analyzed using the FACSCalibur cytometer from Becton Dickinson. Data analysis was performed using Becton Dickinson FACSDiva software.
Measurement of intracellular Ca2+ in synovial cells.
Cells were loaded with 4.4 μM Fluo-3 AM in 0.17% pluronic acid F-17 in an HBSS–1% BSA solution (Sigma) containing 2.5 mM probenecid (pH 7.45) (Sigma) for 30 minutes. After the incubation period, the plates were washed twice with the HBSS–BSA–probenecid solution and placed at 37°C for 30 minutes. Fluorescence was measured at an excitation wavelength of 460–490 nm and emission wavelength of 515 nm.
Calcium influx assays were performed in an automated plate reader (NOVOstar; BMG Labtech). Tested substances were injected into the chamber (25 μl into 75 μl) and the assay was conducted using a kinetic time course of 65 recordings (1 per second). Cells were exposed to tested substances from 10 seconds to 65 seconds. Results are expressed as the ΔF/F, representing the ratio of the fluorescence measurement of each recording to the baseline measurement (mean of the fluorescence intensity between 0 and 10 seconds). Cells were challenged with increasing concentrations of PAR-2–AP (3–1,000 μM), 4α-PDD (3–100 μM), or GSK1016790A (1–30 nM). In further experiments, cells were challenged once with a nearly 50% effective concentration (EC50) of 4α-PDD of 50 μM, after a 5-minute pretreatment with PAR-2–AP or PAR-2–RP at a concentration of 100 μM.
Calcium imaging in neurons.
Neurons were loaded with Fluo-3 AM as described above, and imaged using an inverted microscope (Cell Observer; Carl Zeiss) with a 10 × 0.5–NA objective. Images were acquired using a CCD camera (Carl Zeiss) and Metamorph software. A kinetic setting of 65 recordings (1 per second) was used. From 10 seconds to 65 seconds, neurons were exposed to the different tested molecules. Regions of interest (ROIs) were fitted around the perimeter of a single neuron, and intensity variations for each ROI were measured using Metamorph software. Results are expressed as the ΔF/F. The experiments were repeated 3 times per condition. In the first set of experiments, neurons were exposed to 4α-PDD (50 μM), GSK1016790A (30 nM), PAR-2–AP (SLIGRL-NH2; 100 μM), or their respective controls. In the second set of experiments, neurons were pretreated with PAR-2–AP (100 μM) or the control peptide PAR-2–RP (LRGILS-NH2; 100 μM) for 5 minutes, and were then incubated with 4α-PDD (50 μM).
Assessment of inflammation and pain.
Intraarticular injection of substances into the left (ipsilateral) TMJ was performed in a final volume of 10 μl, as previously described (23). The amount of plasma extravasation was estimated by the extravascular accumulation of intravenously injected Evans blue dye (50 mg/kg) 1 hour prior to the end of the experimental period. The soft tissues of the TMJ were dissected, weighed, and incubated over 24 hours at 60°C in 0.5 ml formamide. The amount of dye loaded in each animal was normalized by measuring the concentration of dye in the plasma at an absorbance of 620 nm. The concentration of dye was estimated by referring to the absorbance obtained with a standard curve of Evans blue dye (1–50 μg/ml), and the values of plasma extravasation (expressed as μl of plasma per gm of TMJ) were determined in accordance with the amount of dye present in 10 μl of plasma and the corresponding TMJ weight. Myeloperoxidase (MPO) activity was assessed in the soft tissues of the TMJs after dissection and freezing of the tissue in dry ice, and was used as an index of granulocyte infiltration, as previously described (23). The results are expressed as units of MPO per joint.
Mechanical allodynia in the TMJ was evaluated using a previously described technique (23, 24). Briefly, animals were acclimatized in the room where tests were performed for 2 consecutive days. On the third day, the basal force threshold value (in gm) was recorded before the intraarticular injection of agonists or controls. Measurements of force thresholds from the ipsilateral and contralateral TMJs at the 1-hour, 4-hour, and 24-hour time points after intraarticular injection were obtained in triplicate, averaged, and expressed as the percent variation relative to the basal force threshold value.
Data were analyzed using Student's t-test for paired data or one- or two-way analysis of variance followed by Tukey-Kramer or Bonferroni post hoc test for multiple comparisons, as appropriate. P values less than 0.05 were considered statistically significant.
TRPV-4 expression in the TMJs and TG neurons.
TRPV-4 immunoreactivity was widely observed in the rat TMJs, with intense labeling in the synovium and condylar cartilage (Figures 1A–C), particularly in the lining layer of the synovial membrane and in chondrocytes of the condylar cartilage. Less intensively, TRPV-4 immunoreactivity was observed in the cells of the sublining layer of the synovium and in blood vessels, both in endothelial cells and in smooth muscle cells. TRPV-4 immunoreactivity in the synovial membrane lining layer colocalized with CD68-positive cells, a tissue macrophage marker, but also colocalized with CD68-negative cells (results not shown).
Intense TRPV-4 immunoreactivity was observed in slices from the TG, colocalizing with the neuronal marker PGP9.5 (Figures 1D–F). Indeed, ∼82% of neurons labeled with PGP9.5 exhibited TRPV-4 immunoreactivity, in populations of neurons varying in diameter (small, medium, or large). Among a population of 294 cells, the mean ± SEM diameter of the nucleated perikarya of TRPV-4–immunoreactive neurons was 30.5 ± 0.5 μm (Figure 1G).
Signaling of TRPV-4 agonists in fibroblast-like synoviocytes and TG neurons.
Flow cytometry analysis showed that only 0.8% of synovial cells from the TMJ were positive for the macrophage marker CD68 at 7 days after dissociation (Figure 2A), as observed by confocal microscopy, with only a few fields containing scattered positive cells. Furthermore, CD68-positive cells were characterized by heterochromatin-enriched nuclei, whereas CD68-negative cells presented a larger, euchromatin-rich nucleus, a feature of fibroblast-like synovial cells. Nevertheless, TRPV-4 immunoreactivity was evident in virtually all of the TMJ cultured synovial cells (Figures 2B–D).
Since the majority of synovial cells had a fibroblast-like phenotype, we then evaluated whether those cells expressed functional TRPV-4. Indeed, the TRPV-4 agonists 4α-PDD and GSK1016790A induced a concentration-dependent (1–100 μM and 1–30 nM, respectively) increase in Ca2+ responses (Figures 2E and F). The EC50 of this response was 33 μM for 4α-PDD and 7 nM for GSK1016790A (Figure 2F).
Acutely dissociated TG neurons expressed TRPV-4 within the neuronal perikarya, but also on neurites (Figures 3A–C). In addition, exposure of TG neurons to 4α-PDD (50 μM) or GSK1016790A (30 nM) induced Ca2+ responses, while exposure to vehicle had no effect (Figure 3D).
Induction of inflammation and mechanical allodynia by TRPV-4 agonists in the TMJs.
The intraarticular injection of the TMJs with the TRPV-4 agonist 4α-PDD (0.07–2 μg) or GSK1016790A (0.02–0.2 μg) induced a dose-dependent increase in plasma extravasation 45 minutes after injection, whereas administration of vehicle had no effect (Figure 4A). Similarly, intraarticular administration of 4α-PDD (0.2–2 μg) or GSK1016790A (0.002–0.2 μg) induced a dose-dependent increase in MPO activity after 4 hours (Figure 4B), and resulted in a time-dependent decrease in the head-withdrawal threshold (Figure 4C), a measure of allodynia.
Coexpression of TRPV-4 and PAR-2 in the regulation of inflammation.
Similar to the findings with regard to TRPV-4, PAR-2 immunoreactivity was observed in the rat TMJs, with an intense labeling in the synovium (Figures 5A–D), especially in the lining layer of the synovial membrane and in chondrocytes of the condylar cartilage (results not shown). Colocalization of TRPV-4 and PAR-2 immunoreactivity in the lining layer was also observed in CD68-positive cells (results not shown). Upon induction of inflammation with carrageenan, the fluorescence intensity of TRPV-4 and PAR-2 immunoreactivity increased 4 hours after intraarticular injection of carrageenan. Strong PAR-2 and TRPV-4 colocalization was observed in the lining layer of the synovial membrane from inflamed TMJs, when compared to vehicle (saline)–treated controls (Figures 5E–H).
In the TG, a high percentage of neurons coexpressing TRPV-4 and PAR-2 was observed. Under physiologic conditions, analysis of a population of 246 neurons showed that 77% of the cells were positive for TRPV-4, 66% for PAR-2, and 58% for both receptors. In addition, among the PAR-2–positive neurons, 88% were also positive for TRPV-4. The neuron found to coexpress both receptors had a diameter that ranged from small to medium, with a mean ± SEM diameter of 29.1 ± 0.6 μm, while neurons expressing only TRPV-4 were, on average, larger in diameter (Figures 5I–L).
Based on the results of real-time RT-PCR and immunofluorescence analyses, we observed that the relative messenger RNA (mRNA) levels (Figures 5M and N) and protein levels of TRPV-4 and PAR-2 were up-regulated at 4 hours after carrageenan injection in the rat TMJs. In the TG, only TRPV-4 was up-regulated at the mRNA and protein levels (Figures 5O and P).
Sensitization of TRPV-4 by PAR-2 activation.
Treatment with PAR-2–AP (3–300 μM) triggered concentration-dependent Ca2+ responses in fibroblast-like synovial cells, with an EC50 of 63.8 μM and a maximal response obtained at a PAR-2–AP concentration of 300 μM (results not shown). We further evaluated whether TRPV-4 could be sensitized by preexposure of the cells to a PAR-2 agonist. Interestingly, pretreatment of the cells for 5 minutes with PAR-2–AP (100 μM) increased the Ca2+ responses to 4α-PDD (50 μM), when compared to the effects of pretreatment with the control PAR-2–RP at the same concentration (Figure 6A). Similarly, pretreatment of TG neurons with PAR-2–AP (100 μM) induced Ca2+ responses that were higher than those after pretreatment with PAR-2–RP, and after 5 minutes of treatment of the neurons with PAR-2–AP (100 μM), the Ca2+ responses to 4α-PDD (50 μM) were potentiated, when compared to that with PAR-2–RP (Figure 6B).
In vivo, the coinjection of low doses of PAR-2–AP (1 μg) and 4α-PDD (0.07 μg) induced plasma extravasation, as shown in Figure 6C. In addition, the intraarticular coinjection of low doses of PAR-2–AP (10 μg) and 4α-PDD (0.3 μg) induced an increase in MPO activity, when compared to that in the respective control groups (Figure 6D). The injection of a noninflammatory dose of PAR-2–AP (1 μg) induced a slight, but significant, decrease in the head-withdrawal threshold. In animals coinjected with 10 μg PAR-2–AP and a noneffective dose (0.07 μg) of 4α-PDD, there was a significant and persistent decrease in the head-withdrawal threshold. This effect was not observed in animals coinjected with PAR-2–RP (Figure 6E).
Among several inflammatory diseases affecting the joints, such as the TMJs, the most relevant are highly debilitating disorders associated with inflammatory pain, such as osteoarthritis and rheumatoid arthritis. While the mechanisms leading to the onset of these diseases are thought to be different, their final stage is associated with an irreversible loss of joint cartilage, in which inflammation of the synovial membrane and locally released proteases have a pivotal role (25). In spite of improvements in the differential diagnosis of these often symptom-overlapping chronic inflammatory diseases, and although advances in treatments have been helpful in achieving long-term management of disease, the fundamental pathophysiologic mechanisms remain poorly understood (26).
Herein, we have identified a putative new mediator of joint inflammation and pain. Indeed, in vivo activation of TRPV-4 by the selective agonists 4α-PDD or GSK1016790A induced a dose-dependent increase in plasma extravasation, granulocyte recruitment, and mechanical allodynia in the TMJs. Furthermore, GSK1016790A induced a potent increase in MPO activity, almost as efficient as the induction of inflammation by carrageenan, suggesting that this agonist induces an important neutrophil chemoattraction signal.
TRPV-4 was recently implicated as a molecular mechanotransducer in chondrocytes, with results showing that this channel could mediate Ca2+ influx in response to osmotic changes or 4α-PDD (10, 27). These findings have important outcomes, considering that chondrocytes can be exposed to hypo-osmolarity during joint loading or degeneration of the extracellular matrix of cartilage during inflammatory arthritis (28, 29). Indeed, deletion of the Trpv4 gene in mice resulted in severe osteoarthritic changes (10). Moreover, single-point mutations in TRPV-4 reduced its transport to the correct subcellular domain of the cell plasma membrane and disrupted its response to hypo-osmotic stimulation, leading to an inherited arthropathy in the fingers and toes (11), suggesting that this channel plays a chondroprotective role. Intriguingly, mutations in the Trpv4 gene encoding for a channel with gain-of-function have also been linked to several autosomal-dominant musculoskeletal dysplasias (9, 30–33).
Overall, these findings suggest a role for TRPV-4 in the homeostasis of the musculoskeletal system, including the joints. A tight regulation of its activity seems mandatory, since either loss-of-function or gain-of-function of TRPV-4 is detrimental to the development of several disorders affecting this system. In this way, either the lack of activity or the overactivity of TRPV-4 in chondrocytes may impair proper cell adaptation to osmotic changes, with implications for cell viability and/or chondrogenic activity.
In the present study, we showed that TRPV-4 was mostly expressed in the lining layer, both in macrophage-like (CD68-positive) cells and fibroblast-like cells in the rat TMJ synovial membrane. Human primary synoviocytes and the synovial cell line SW982 are known to express functional TRPV-4 (34). Moreover, the stimulation of this cell lineage with tumor necrosis factor increased the expression of TRPV-4 and its response to thermal or osmotic stimulation (32). Herein, the stimulation of cultured fibroblast-like synoviocytes with 2 selective TRPV-4 agonists induced Ca2+ responses, further confirming the functionality of TRPV-4 in these cells.
Thus far, the role of TRPV-4 in these cells remains unknown, but PAR-2 signaling seems to be a pivotal mediator of fibroblast-like synoviocyte transformation, leading to uncontrolled cell proliferation and invasiveness of joint subchondral bone (33). TRPV-4–mediated Ca2+ responses were potentiated by prestimulating fibroblast-like synovial cells and TG neurons with PAR-2–AP. Likewise, the coinjection of subinflammatory doses of TRPV-4 and PAR-2 agonists into the TMJ induced the development of inflammation and allodynia. Taken together, these findings suggest that the interplay between TRPV-4 and PAR-2 could take place in joint cells expressing both receptors during inflammation.
The occurrence of this functional interaction endogenously and its physiopathologic outcome in human inflammatory diseases still remain to be elucidated. Nevertheless, in this study, we were able to determine the molecular substrates for such an interaction. TRPV-4 and PAR-2 were highly coexpressed in the synovial lining layer of the TMJs, in chondrocytes, and in TG neurons. Furthermore, a significant increase in TRPV-4 and PAR-2 expression, either at the transcriptional level or at the translational level, was evident in the TMJs, and a significant increase in TRPV-4 was found in the TG 4 hours after carrageenan-induced inflammation, a time point at which a characteristic pattern of marked plasma extravasation, leukocyte recruitment, and pain has been observed (23).
PAR-2 is known to be up-regulated in chondrocytes and synoviocytes during osteoarthritis and rheumatoid arthritis, and deletion of its gene substantially reduces the development of arthritis in mice (35–39). Furthermore, the number of mast cells and release of the PAR-2 agonist tryptase are increased during arthritis (40, 41) and in other inflammatory diseases such as irritable bowel syndrome (42). In an arthritis model highly dependent on the human tryptase mouse ortholog mMCP6 (mouse mast cell proteinase 6), synovial macrophages supported the final differentiation of Th17 CD4+ T cells by producing large amounts of interleukin-1β (IL-1β), IL-6, IL-23p19, and transforming growth factor β (41). Moreover, although TRPV-4 can be directly activated by mechanical stretch, moderate heat, 4α-PDD, or GSK1016790A (1, 43), its response to hypo-osmolarity is dependent on 5′,6′-epoxyeicosatrienoic acid formation by cytochrome p450 epoxygenase from arachidonic acid (3), which is abundant during joint inflammation (44, 45).
In TG neurons, TRPV-4 immunoreactivity was observed at the plasma membrane and in the nuclear/perinuclear regions of perikarya. The intracellular localization of TRPV-4 has already been described in dorsal root ganglia neurons, in a study showing that pools of this receptor were ready to be sent to the plasma membrane upon stimulation with inflammatory mediators (17). In a previous study, we have shown that 84% of TMJ primary sensory afferents expressed PAR-2 (23). Considering that almost all of the PAR-2–positive neurons were also positive for TRPV-4 (88%), we can infer that a large population of TG neurons that project to the TMJ will coexpress both receptors.
It was recently shown that TRPV-4 activation induced inflammation in the mouse paw by neurogenic mechanisms (21). Considering that a functional interaction of TRPV-4 and PAR-2 can take place in neurons, as has been shown by in vitro and in vivo experiments, and based on previous findings that PAR-2–induced inflammation and pain are dependent on calcitonin gene-related peptide and substance P (19, 46, 47), the antidromic release of these neuropeptides is a major candidate mechanism for the induction of the inflammatory response after the coactivation of TRPV-4 and PAR-2 in the TMJs. This hypothesis is further strengthened by previous studies showing that PAR-2 activation in neurons is a strong signal that can sensitize TRPV-4, thus implicating this receptor as an effector channel in the processes of nociception and neurogenic inflammation induced by serine proteinases (16, 20, 48).
Taken together, these results show that TRPV-4 is expressed in cells with recognized roles in the physiopathology of arthritis, and that its activation induces joint inflammation and pain. Furthermore, TRPV-4 is highly coexpressed with PAR-2 in these cells, and an interaction between both receptors potentiates inflammatory responses. Although the endogenous activators of both receptors in the synovial joints remain to be better characterized, and although it is yet to be determined whether this interaction can take place in human joint disease, our study shows, for the first time, that these receptors may account for a multicellular system with a putative role in the onset or the severity of inflammatory arthritis.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Vergnolle had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Denadai-Souza, Werneck de Avellar, Muscará, Vergnolle, Cenac.
Acquisition of data. Denadai-Souza, Martin, Vieira de Paula, Cenac.
Analysis and interpretation of data. Denadai-Souza, Martin, Vieira de Paula, Vergnolle, Cenac.
We thank S. A. Teixeira, I. M. Gouvea, M. A. A. G. Barreto, C. Z. Romera, E. N. Kanashiro, and E. M. de J. S. Santos (São Paulo, Brazil) for providing valuable technical help, and Sophie Allart (Toulouse, France) for providing technical assistance at the cellular imaging facility of INSERM 1043.