Background
Temperature sensitive transient receptor potential (TRP) channels belonging to the V- (or vanilloid related) subfamily are widely expressed in mammalian cells. Four members of this subfamily, TRPV1–4 conduct mono- and di-valent cations when activated by temperatures ranging from > 23°C (TRPV3 and TRPV4) to > 43°C (TRPV1) or > 53°C (TRPV2). In addition, TRPV1–4 function as important membrane sensors for extracellular chemical, osmotic, or mechanical stimuli. TRPV1 channels are activated by low pH (< 5.9) and endovanilloids. TRPV4 and TRPV2 respond to cellular swelling and mechanical stimulation. TRPV1 and TRPV4 are activated by anandamide and arachidonic acid metabolites [
1‐
4]. Other TRP family channels also respond to chemical (icilin and camphor, TRPA1 and TRPV3) and menthol or cold stimulation (TRPM8, < 19°C). This ligand promiscuity supports an important role of TRP channels during episodes of acute or chronic inflammation, where dramatic changes in the extracellular environment impact the physiological and chemical homeostasis.
Several recent studies have demonstrated that neuronal growth factors and proinflammatory chemokines and cytokines can increase the physiologic response of TRP channels [
5‐
8] and emphasized the enhancing role of TRP channels in chronic inflammation [
3,
8‐
10]. In a previous study [
11] we showed functional expression of thermal and chemical sensitive (TRPV1, TRPV4, TRPA1) TRP channels on SW982 clonal and primary human synovial cells. We have proposed a significant role for TRP in mediation of calcium dependent proliferative and secretory responses of synoviocytes during joint inflammation. The present study supports our hypothesis demonstrating increases in thermal and osmotic sensitive TRP channel mediated responses after exposure to proinflammatory modulator tumor necrosis factor alpha (TNF-α).
TNF-α is found in abundance in synovial fluid of patients with arthritis [
12], and the receptors for TNF-α are found on synoviocytes harvested from patient tissue or synovial fluid [
13,
14]. It is reported by Youn and colleagues [
14] that upon TNF-α stimulation of synoviocytes harvested from patients, the cells proliferate and the expression of TNFR2 increases dose dependently, reaching a maximal level after 24 h of stimulation. In contrast, the levels of TNFR1 transcripts decrease up to 12 h after TNF-α stimulation in a time-dependent manner. TNF-α very effectively increases chemokine production in fibroblast-like synoviocytes harvested from arthritis patients [
15]. TNF-α has been shown to be a first line initiator of inflammatory responses in joints since synovial lining cells express TNF-α prior to the appearance of other cytokines in a collagen induced arthritis model [
16].
Discussion
In the present study we demonstrated up-regulation of two members of the TRP cation channel family in SW982 human synovial cells after preincubation with TNF-α. The TNF-α induced increases in TRPV1 and TRPV4 gene and protein expression at 8 and 12 hr respectively, are coincident with significant enhancement of responses to increasing temperature and decreasing osmolarity mediated by TRPV1 and TRPV4. These two TRP channels appear to be the major thermal and osmosensors in synoviocytes. Marginal facilitation of the calcium response at 30°C, in synoviocytes incubated with 2-AB demonstrates the potential for TRPV3 mediated response to thermal stimuli in some circumstances, although no response to camphor was generated. TNF-α did not induce a change in synoviocyte responsiveness to TRPA1 (icilin) or TRPM8 (menthol) activation.
A number of endogenous factors can activate TRPV1 and TRPV4 during inflammation (for review see [
3,
10]). This includes factors generated in vast amounts during inflammation, i.e. direct ligands such as protons, temperature increases, lipid mediators and hypoosmolarity, or indirect activators such as cytokines, bradykinin, ATP, serotonin or PGE2 [
8]. We have previously shown responsiveness of synoviocytes to capsaicin, resiniferatoxin and icilin [
11]. The increased levels of extra-physiologically activated TRPV1 and TRPV4 channels caused by incubation with TNF-α can significantly contribute to the activation of synoviocytes observed in joint inflammation. Like other fibroblasts, the fibroblast-like synoviocytes secrete TNF-α [
20], as well as express its receptors [
13]. It is highly likely that TNF-α increases expression of TRPV1 and TRPV4 by mediating the up-regulation directly and/or through initiation of downstream events since the inhibitor reduced the immunostaining increase. Further studies are required, however.
Sensitization of TRPV1 and TRPV4 was demonstrated after TNF-α exposures with increased cell responsiveness to chemical agonist capsaicin, noxious thermal and hypoosmolar stimulation. It is still obscure, why only a fraction of the cultured clonal synoviocytes had increased responsiveness to capsaicin, as all cells demonstrated increased [Ca
2+]
i in response to noxious temperature and icilin. It is also unclear why the synoviocytes readily responded to hypoosmolarity but not to TRPV4 agonist 4α PDD. This heterogeneous response suggests that these TRP channels exist in multi-potential functional states, that non-neural cellular responses are not identical to neuronal responses, and/or that tertiary protein structure plays a role in activation properties. Other mechanisms regulating TRP channel sensitization may be involved, such as PKC or PKA dependent phosphorylation [
8,
9,
21‐
24] or phosphatidylinositol phosphate binding [
1,
25]. Further studies detailing the molecular mechanisms of TNF-α induced increase of TRPV1 and TRPV4 mediated response in synoviocytes are needed.
Specificity of the TNF-α induced sensitization for TRPV1 and TRPV4 responses was demonstrated. No significant change in calcium flux from baseline and numbers of synoviocytes responding was noted after TNF-α pre-treatment (12 hr) for menthol, a TRPM8 activator or camphor, a TRPV3 activator. While almost all of the cells responded to icilin, TNF-α pre-treatment did not alter the response properties of the synoviocytes to the TRPA1 stimulation. Lack of response to camphor and lack of sensitization of the menthol response provide evidence in agreement with Kochukov et al. [
11] that TRPV3 and TRPM8 are minimally expressed in this human SW982 synoviocytes cell line. Thus, TNF-α sensitized thermal and hypoosmotic responses of the synoviocytes are mediated primarily by TRPV1 and TRPV4 activation respectively.
The present study was performed on the human SW982 synovial sarcoma cell line, which a number of studies have shown share similar physiological properties, cytokine expression and regulation profiles with primary human type B synovial cells. Our previous data demonstrate the functional expression of TRPV1, TRPV4 and other TRP channels in both SW982 cells and primary human synoviocytes [
11]. In all aspects the results were similar for the clonal and primary cells. The presence of functional TRPV1 in human synovial and dermal fibroblasts has been confirmed by other authors [
26,
27]. In a separate study we also demonstrated close similarity between primary human synoviocytes and SW982 cells in the functional expression of a G-protein coupled membrane acid sensing receptor [
28].
Many previous studies describing TRP channel up-regulation have been performed on neuronal cells, where TRPV1 and/or TRPV4 sensitization are shown to be an important mechanism contributing to chronic hyperalgesia [
8,
29‐
33], bronchial hyper-responsiveness [
34], neurogenic bladder [
35], and neurogenic inflammation [
9,
36‐
39] (for detailed review [
3,
10]). In contrast, TRP channel function and regulation in non-neuronal cells such as synovial fibroblasts are less well studied, although the physiologic contribution of TRPV1 in experimental arthritis has been well documented recently using TRPV1-knockout mice [
8,
31,
40‐
42]. Abundant expression of TRPV1 in human synovial cells and its up-regulation in the membrane fraction with prolonged TNF-α exposure support our hypothesis that non-neuronal TRP channels are participants in development of peripheral inflammation.
Our current study demonstrates increased TRPV4 mediated calcium mobilization in response to hypotonicity in synoviocytes. However, we did not initially observe significant changes in hypotonicity-induced calcium response in synoviocytes after TNF-α co-stimulation at 2, 4, and 8 hr, despite the moderate temperature sensitization. Upon further study, however, the functional consequences of TRPV4 up-regulation by TNF-α become evident after 12 hr and persisted through 16 hr of study. The average amplitude of the TRPV4 activated calcium responses was significantly elevated and was blocked by ruthenium red. The numbers of cells responding to hypoosmotic stimuli were greatly increased 3-fold at 12 hr and 4.5-fold at 16 hr.
The role of TRPV4 in synoviocytes is less clear as is the delay in activation response. Edema is a cardinal sign of inflammation and osmolarity changes may be a factor. The delayed TRPV4 sensitization at 12 hr was coincident with increased TRPV4 immunostaining and protein content, and development of a flattened appearance of the cells. These increases became evident when TRPV1 staining had returned to normal levels. A universal role for TRPV4 in osmosensation and mechanotransduction has been promoted in recent studies based on emerging data from neural cells, epithelia, mesenchymal cells, kidney and osteoclasts ([
2], reviewed in [
43]). It is noted in neurons that simultaneous action of a number of combined inflammatory mediators is required to achieve sufficient activation of the cAMP pathway to allow TRP-dependent hyperalgesia to occur [
8]. The present study demonstrates that the inflammatory mediator TNF-α which releases a cascade of other mediators and enzymes is sufficient over time to promote the shift in synoviocytes to osmoreceptor mediated through TRPV4.
Surprisingly, the TRPV4 response to 4α PDD was absent in the human synoviocytes. This difference may be explained by distinctive properties of synoviocyte TRPV4 as heterogeneity of this channel due to alternative splicing and heteromultimerization has been reported [
44]. The TRPV4 transmembrane segment TM3 is a crucial region for the activation by its most specific agonist 4 alpha-phorbol 12, 13-didecanoate (4α-PDD). Vriens et al. [
45] have found that mutation of one tyrosine (Tyr555) at the intracellular N terminal of TM3 (S3) leads to a loss of responsiveness to 4α-PDD, despite the demonstrated presence of a functional channel activated by hypotonic stimulation. This suggests inability of 4α-PDD binding at the single critical tyrosine in the TM3 segment in human synoviocytes likely due to the tertiary folding of the protein in this region as suggested previously [
45].
The present data indicate that synoviocytes express functional thermal and osmotic receptors inducible under inflammatory conditions. The TNF-α mediates up-regulation of TRPV1 and TRPV4, doubling mRNA and protein expression at 8 and 12 hr, respectively. Microarray data has been published detailing downstream targets for TNF-α in synoviocytes [
46] and demonstrates regulation of numerous genes by multiple pathways. Increased expression of vanilloid channels in the membrane fraction in the present study implies translocalization of TRPV1 and TRPV4 to the plasma membrane of synoviocytes initiated in response to the TNF-α. The time course of TNF-α induced increase in TRPV1 and TRPV4 protein expression correlates well with the observed TNF-α induced increases in thermal and hypoosmotic induced [Ca
2+]
i responses and cell recruitment.
The literature suggests that TRP expression in epithelial lining cells may be extremely important in cells that interface with dramatic changes in the external and internal environment induced by thermal and mechanical stress, acidity, hypoosmolarity, foreign chemicals and endogenous mediators. Synoviocytes are exposed to all of the above stimuli in the course of routine joint function and in acute and chronic inflammatory states. The synoviocytes are like other fibroblasts that secrete the inflammatory mediator, TNF-α [
20], as well as express TNF-α receptors [
13,
14]. It is known that the TNF-α can directly activate TNF-α receptors on peripheral nerve terminals to amplify hyperalgesic responses [
47,
48] and release enzymes causing injurious joint destruction and impaired biomechanical function [
49‐
51]. The TRPV1 and TRPV4 mediated [Ca
2+]
i signaling responses of synoviocytes during inflammation would participate in the TNF-α mediated functions that promote the painful inflammatory and joint destruction processes locally and impact subsequent responses of the immune system. The data suggest TRPV1 and TRPV4 play important roles mediating the intracellular Ca
2+ increase that drives critical functions such as associated proliferative and secretory responses of synoviocytes during joint inflammation thus promoting TRPV1 and TRPV4 as attractive potential therapeutic targets.
Methods
Cell lines and cultures
The fibroblast-like synovial SW982 [
42] cells were obtained from the American Type Culture Collection (ATCC, Bethesda, MD) and maintained in Leibovitz's L-15 medium (GIBCO, Grand Island, NY), 10% heated fetal bovine serum (FBS, Gemini Bio-Products, Woodland, CA), 2 mM L-glutamine, 1,000 U/ml penicillin G (GIBCO), and 100 ng/ml streptomycin (GIBCO). Cells were incubated in a humidified cell incubator (37°C, ambient CO
2). Low passage cells (3–15×) were used in this study.
Cytosolic free calcium ([Ca2+]i) measurement
For measurement of [Ca
2+]
i in cells exposed to chemical activators of different TRP channels a fluorescent calcium imaging approach was used. Cells were harvested with 0.25% trypsin-0.02% EDTA disruption, plated on 15-mm circular quartz glass coverslips at a density of 200,000 cells/mm
3 and incubated at 37°C for 24–48 h before an experiment. On the day of the experiment, the cells were loaded with the calcium-sensitive fluorescent dye fura-2 as previously described [
11]. Fluorescent recording of temperature-induced calcium changes was performed as described earlier [
11] using Nikon Diaphot microscope with an 60× water immersion lens and Till Photonics Polychrome II photometry setup (Munich, Germany) equipped with Hamamatsu R928 photomultiplier and CL-100 bipolar temperature controller with a SC-20 inline solution heater/cooler (Warner Instruments, Hamden, CT) and controlled by ITC-18 computer interface (Instrutech, Port Washington, NY) and X-chart software(HEKA, Heidelberg, Germany). Experiments were repeated a minimum of four times each. The 340-to-380 ratios acquired from single or two-three neighboring cells every 200 ms were converted into [Ca
2+]
i with the formula [Ca
2+]
i =
K
d [(R-R
min)/(R
max - R)](S
f2/S
b2)[
52], in which fluorescence ratio sat zero free Ca
2+ (R
min) and saturation free Ca
2+ (R
max), as well as fluorescent intensity of Ca
2+-free (S
f2) and Ca
2+-bound(S
b2) dye, excited at 380 nm, were measured experimentally and fura-2
K
d values for Ca
2+ at different temperatures were calculated in accordance Shuttleworth and Thompson estimations [
53]. Each calcium trace was evaluated individually and the response was considered positive when the elevation in [Ca
2+]
i was clearly seen at the appropriate time and when the [Ca
2+]
i returned to baseline in the absence of treatment. Detection threshold of [Ca
2+]
i changes was dictated by signal to noise ratio and was at least 10 nM over the baseline. Peak and average amplitude of [Ca
2+]
i changes over a baseline were used for quantitative estimation of calcium response.
For simultaneous measurement of [Ca2+]i in large numbers of cells exposed to chemical activators of different TRP channels, human SW982 synoviocytes were plated onto 12 mm coverslips, cultured for 24 hr, then pre-treated with TNF-α (1 ng/ml). The cells were loaded with the calcium-sensitive fluorescent dye fura-2 for 30–60 min before calcium recording. The activation was done on 3 plates with over 60 cells selected. The imaging setup included a Nikon 200E microscope with a 20× SuperFluo lens, and a computer-controlled illumination system (Sutter Instruments, Novato, CA) equipped with a digital monochrome-cooled charge-coupled device RoperCoolsnap HQ camera (Roper Scientific, Tucson, AZ). Fluorescent emission at 510 nm from regions of interest (corresponding to a single cell) was acquired online with the MetaFluor software (Universal Imaging, Downington, PA). The signal was obtained in dual, 340 and 380 nm, excitation mode and average intensity of fluorescence in each region after were used to estimate 340-to-380 ratios. The elevation in calcium influx for the individual and population data was considered an increased response if statistically different from the baseline.
Chemicals
The fluorescent intracellular calcium release indicator fura-2AM was obtained from Molecular Probes (Eugene, OR). TRPV1 agonist (capsaicin) and antagonists (capsazepine and 2-aminoethoxydiphenyl borate (2-AB; also antagonizes TRPV3 but not TRPV4, inhibits TRPM8), were obtained from Tocris (Ellisville, MO), as was the icilin (TRPA1 agonist). The ruthenium red (RR, TRPV1 and TRPV4 antagonist), 4-α-phorbol 12, 13-didecanoate (4α-PDD, TRPV4 antagonist), and camphor (TRPV3 agonist) were obtained from Sigma (St. Louis, MO). L-menthol (TRPM8 agonist) was obtainedfrom Aldrich (Milwaukee, WI). The inflammatory cytokine tumor necrosis factor alpha (TNF-α) was obtained from Calbiochem (San Diego, CA). The TNF-α inhibitor CAY10500 (6,7-dimethyl-3-{[methyl-[1-(3-trifluoromethyl-phenyl)-1 H-indol-3-ylmethyl]-amino}-ethyl)-amino]-methylchromen-4-one) was obtained from Cayman Biologicals (Ann Arbor, MI). Thapsigargin used to test cytosolic calcium release at the end of each study was obtained from Calbiochem (La Jolla, CA).
Real time PCR
The low passage SW982 cell cultures were grown in 75 mm flasks until they reached 70–80% confluency. Leibovitz's L-15 Medium (Invitrogen, Carlsbad, CA) with 10% FBS, 1% P/S was used for culture medium and the cells were incubated at 37°C, room air, per ATCC recommendations. For the experiments, TNF-α (1 ng/ml) was added to the designated flasks. The cells were harvested by disruption (see below) over a time course of baseline (immediately after TNF-α addition), and at 4 and 8 hours after addition. Flasks containing untreated cells were also harvested at the same time as baseline controls. These studies were performed in three separate experiments.
Total RNA was isolated using Trizol reagent (Cat# 10296028, Invitrogen) in accordance with manufacturer's protocol. First-strand cDNAs were synthesized from total RNA by RT using the SuperScript® III First-Strand kit (Cat# 18080051, Invitrogen) with random hexamer primers as described by the manufacturer. Two microgram of total RNA from each individual sample was used to synthesize cDNA for TRPV1, TRPV4 and β-actin, respectively.
Quantitative real-time PCR analysis was performed by the use of an Applied Biosystems 7000 Sequence Detection System (Applied Biosystem Inc, Foster City, CA). Each sample was analyzed in triplicate and the means were used for statistical purposes. The thermal profiles were obtained by using 2 min incubation at 50°C, followed by an initial 10 min denaturation step at 95°C, and by 40 cycles of 1 min each at 60°C plus 15 sec at 95°C. Significant contamination with genomic DNA was excluded by amplifying non-reverse-transcribed RNA.
The probes and primers for TRPV1 (
Hs00218912_m1), TRPV4 (
Hs00540967_m1) and β-Actin (
Hs99999903-m1) were ordered from Applied Biosystem Inc. The threshold cycle (Ct), i.e. the cycle number at which the amount of amplified gene of interest reached a fixed threshold, was determined subsequently. Relative TRPV1 and TRPV4 mRNA expression was calculated by the comparative Ct method described elsewhere ()[
54]. Analysis of relative gene expression data using real time quantitative PCR and the 2-Ct method. The relative quantitative value of target, normalized to an endogenous control β-actin gene and relative to a control, is expressed as 2
-ΔΔCt (fold), where ΔCt = Ct of target gene (TRPV1 or TRPV4)-Ct of endogenous control gene (β-actin), and ΔΔCt = ΔCt of samples for target gene-ΔCt of the control for the target gene. 2 μl of synthesized cDNA from each individual sample was used to amplify for TRPV1, TRPV4 and β-actin, respectively.
Western blot analysis
Low passage SW982 cell cultures were grown in 75 mm flasks until they reached 70–80% confluency. Proteins were extracted from control cultures without treatment and after treatment with 1 ng/ml TNF-α (Pierce Biotechnology, Inc., Rockford, IL) for 8 hours. Western Blot analysis was performed on whole cell protein and subcellular protein fractions extracted from synovial SW982 cells with ProteoExtract Subcellular Proteome Extraction kit (Calbiochem). Protein transferred to membranes were probed with anti-TRPV1 antibody (VR1, P-19, 1:1000, Santa Cruz, Burlingame, CA) and TRPV4 (VR4, K-18, 1:200, Santa Cruz) antibody, at 4°C overnight. Amounts of total protein loaded per lane: 15 μg for cytosolic and membrane fractions, 10–15 μg for whole cell and 5 μg for rat spinal cord lysate as a positive control. The protein analysis was performed with protein from three experiments.
Immunocytochemistry
Cells on glass coverslips were fixed with 4% freshly mixed paraformaldehyde:methanol (5:1, v/v) for 30 min at room temperature. The coverslips were washed 3× in PBS and incubated at room temperature for 30 minutes with 3% neutral serum. The primary antibody (anti-TRPV1, (1:500, P19: sc-12498, Santa Cruz, Burlingame, CA) or TRPV4 (1:400, #ACC-034, Alomone, Jerusalem, Israel) was diluted in TBS with 1.0% neutral serum (0.02% Triton X-100). The primary antibody was allowed to incubate on the glass coverslips for 48–72 h at 4°C. The glass coverslips were then washed 3× with 1% serum, 0.02% Triton X-100 and then incubated for 30–60 min at room temperature with the appropriate secondary IgG antibody (1:200, 1 hr), using donkey anti-goat (for TRPV1) and goat anti-rabbit (for TRPV4). Each had a FITC fluorescent tag (1:200, Santa Cruz). After final washes 3× with PBS, the stained coverslips were inverted onto a dot of VectaShield mounting media containing blue nuclear counterstain DAPI. Immunocytochemical controls included the absence of the primary or secondary antibodies, or were done in the presence of matched serum at the same dilution. The specificity of the antibodies was confirmed for TRPV1 by adsorption controls with P-19 N-terminus peptide provided by Santa Cruz, and for TRPV4 with peptide 853–871 supplied by Alomone Labs (1:1 w/w, 30 min). There was no staining in either the method controls or the peptide blocked controls. TNF-α stimulated increase in TRPV1 was greatly reduced by co-incubation with a TNF-α inhibitor, CAY10500. Stained cells were visualized with a Nikon FXA microscope equipped with MetaVue software (Nikon Instruments, Inc. Melville, NY).
Software and statistics
Sigma Plot and Sigma Stat scientific software (SPSS, Chicago, IL) were used for conversion and analysis of acquired data. Data are reported as means ± SE. MetaMorph software (Universal Imaging, Downington, PA) were used for off-line analysis of calcium imaging data. Student's t-tests or ANOVA for multiple comparisons with Tukey's multiple comparison post hoc test or Dunnett's method were used to compare data obtained from TNF-α treated vs. control series, unless otherwise indicated.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
All authors read and approved the manuscript. MYK was responsible for calcium imaging responses to capsaicin, thermal and osmotic challenges, as well as the initial manuscript draft; TAM was responsible for experimental and primer design for synoviocyte culture studies; HY was responsible for RT-PCR; LZ and FM were responsible for all calcium imaging studies with TNF-α primed synoviocytes; SA and LP were responsible for western blots of whole cell and cell fractions of control and TNF-α primed synoviocytes; SA was responsible for immunolocalization studies, microscopy, cell area and MetaMorph analyses; and KNW was responsible for experimental design of the studies, data synthesis, data figures and manuscript editing for publication.