Defining the cellular repertoire of GPCRs identifies a profibrotic role for the most highly expressed receptor, protease-activated receptor 1, in cardiac fibroblasts

Reagents

All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA)except α-thrombin (Enzyme Research Laboratories, South Bend, IN, USA), angiotensin II (Tocris Bioscience, Ellisville, MO, USA), or as specified below.

Cardiac fibroblast isolation and culture

CFs were isolated from adult male Sprague-Dawley rats as described previously (11). Briefly, adult (8–10 wk) rats (∼200 g) were anesthetized by intraperitoneal injection of ketamine (100 mg/kg) and xylosine (10 mg/kg). The heart was excised and then perfused with medium containing collagenase type 2 (Worthington Biochemical, Lakewood, NJ, USA) via a modified reverse-Langendorff apparatus. The heart was minced and triturated to separate cells from decellularized matrix. The CFs were separated from myocytes by gravity separation, resuspended in DMEM + 10% FBS + penicillin-streptomycin, and plated on 10-cm dishes. On reaching confluency, fibroblasts were passaged into 6- or 12-well plates (150,000 or 70,000 cells/well, respectively) and serum starved for 24 h before treatment. All experiments were done with cells passaged only once, with the exception of the RT-PCR GPCR array, for which unpassaged CFs were used.

RT-PCR GPCR array

Freshly isolated CFs were plated and grown in DMEM + 10% FBS + penicillin-streptomycin until ∼60% confluent. Cells were then serum starved for 24 h, and RNA was isolated using an RNeasy kit with DNase treatment (Qiagen, Valencia, CA, USA). RNA was converted to cDNA using SuperScript III reagent (Invitrogen, Grand Island, NY, USA) using the recommended protocol. GPCR expression was determined with Taqman GPCR Arrays (Applied Biosystems, Life Technologies Corp., Carlsbad, CA, USA) according to the recommended protocol, run on an ABI Prism 7900HT system (Applied Biosystems), and analyzed with the Sequence Detection System software (Applied Biosystems). Quantification of GPCR expression was expressed relative to 18S rRNA. Data shown are from the average of 2 independent arrays.

A table is available from the authors that provides average expression [dC(t)] data (with sd) for all 342 GPCRs (190 detected, 152 not detected), 18 control genes, and 20 GPCR-associated genes, in addition to the array-specific identification number, gene abbreviation, formal gene name, U.S. National Center for Biotechnology Information (NCBI) Reference Sequence (RefSeq), GenBank mRNA ID, and amplicon length for each probe set.

Quantitative RT-PCR

Total RNA was extracted from samples using TRIzol reagent (Invitrogen) according to the manufacturer's protocol and converted to cDNA using SuperScript III reagent. Gene-specific primers designed as described previously (ref. 12;1 μM final concentration) were mixed with cDNAs and SYBR green reagent (Quanta Biosciences, Gaithersburg, MD, USA) for PCR amplification on a DNA Engine Opticon 2 system (Bio-Rad, Hercules, CA, USA). Gene expression levels were quantified relative to that of 18S rRNA. Oligonucleotides used for RT-PCR are shown in Table 1.

Immunoblot analysis

Cell lysates were collected by scraping wells in the presence of 150 mM carbonate buffer (pH 11) and homogenized via sonication. Total protein concentrations were determined using a dye binding reagent (Bio-Rad) and were separated by SDS/PAGE in 10% polyacrylamide gels (Invitrogen). Gels were transferred to PVDF membranes using an iBLOT transfer machine (Invitrogen). Membranes were blocked using 5% nonfat dry milk (w/v) in PBS Tween-20 (PBST) solution and incubated with primary antibody overnight at 4°C. Blots were visualized after binding of secondary antibody conjugated with horseradish peroxidase (Santa Cruz Biotechnology, Santa Cruz, CA, USA) using SuperSignal Westdura Substrate (Thermo Scientific, Rockford, IL, USA). Densitometric quantification of protein expression was done by using ImageJ software (U.S. NIH, Bethesda, MD, USA). All data were normalized relative to β-tubulin. Antibodies against α-smooth muscle actin (αSMA), plasminogen activator inhibitor 1 (PAI-1), and β-tubulin were purchased from Invitrogen, BD Biosciences (San Jose, CA, USA), and Abcam (Cambridge, MA, USA), respectively.

Immunofluorescence microscopy

Immunofluorescence microscopy was performed with slight modification of a previous procedure (11). CFs were plated onto fibronectin-coated coverslips in 24-well plates (70,000 cells/well) overnight, serum starved for 24 h, and then treated as indicated for 24 h at 37°C. The CFs were then washed with PBS and fixed with 10% buffered formalin, permeabilized with 0.3% Triton X in PBS, blocked, and probed with anti-αSMA antibody (Invitrogen). Samples were then washed and incubated with Alexa Fluor 488-conjugated secondary antibody (Invitrogen), washed, and mounted on slides with Vectashield containing 4′,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA, USA). Images were obtained with a Deltavision RT Deconvolution Microscope (Applied Precision, Issaquah, WA, USA) at the University of California at San Diego, Microscope core facility and analyzed with SoftWorx software (Applied Precision) on a Silicon Graphics Octane workstation (Silicon Graphics International, Fremont, CA, USA).

Collagenase-sensitive [³H]proline incorporation

[³H]proline incorporation by rCFs was determined with slight modification of a previously described protocol (11). Briefly, cells were passaged in 12-well plates (70,000 cells/well) in DMEM + 10% FBS + penicillin-streptomycin and grown overnight. Cells were washed, incubated in serum-free DMEM + penicillin-streptomycin for an additional 24 h, and then incubated with test ligands in the presence of [³H]proline (1 μCi/ml final concentration; PerkinElmer, Waltham, MA, USA) for 24 h at 37°C. Following the treatment period, cells were washed with PBS and solubilized with 0.5 N NaOH (300 μl/well) for 5 min at 37°C. Cell lysates were then neutralized with 0.5 N HCl and precipitated with trichloroacetic acid (TCA) overnight (final concentration 20%). TCA-insoluble material was pelleted by centrifugation (>10,000 rpm for 10 min), washed 3× with 5% TCA, and solubilized with 200 μl of 0.2 N NaOH. pH was neutralized with 0.2 N HCl, and samples were incubated with collagenase II (final concentration 2 mg/ml collagenase in Tris-CaCl₂-N-ethylmaleimide buffer) for 1 h at 37°C. After this incubation, undigested material was precipitated with TCA (10% final concentration at 4°C for 1 h) and centrifuged (>10,000 rpm for 10 min). Radioactivity in the supernatants, which contain collagenase-sensitive [³H]proline, representing [³H]proline incorporated into collagen, was quantified using a liquid scintillation counter.

[³H]thymidine incorporation

[³H]thymidine incorporation by rCFs was determined with a slight modification of a previously described protocol (13). Briefly, cells were passaged in 12-well plates (70,000 cells/well) in DMEM + 10% FBS + penicillin-streptomycin and grown overnight. Cells were then washed, incubated in serum-free DMEM + penicillin-streptomycin for an additional 24 h, and then treated with test ligands in the presence of [³H]thymidine (1 μCi/ml final concentration, PerkinElmer) for 24 h at 37°C. Cells were then washed with cold PBS + 7.5% TCA and solubilized with 0.5 N NaOH (5 min at 37°C). Radioactivity in the lysates was quantified using a liquid scintillation counter.

siRNA transfection

CFs were plated in 12-well plates (70,000 cells/well) overnight and were transfected with scrambled negative control siRNA (Ambion, Grand Island, NY, USA) or PAR1-specific siRNA (F2r, s129806; Ambion) at 5 nM final concentration using Lipofectamine RNAi-max (Invitrogen). After 4 h, the medium was replaced with serum-free DMEM; incubations were continued for 24 h before the indicated treatments.

Statistical analysis

All data analysis was performed using GraphPad Prism 4.0 software (GraphPad Software, La Jolla, CA, USA). Data are presented as means ± se. Statistical calculations were done using 1 way ANOVA with either Dunnett or Bonferroni posttests. Values of P ≤ 0.05 were considered significant.

PAR1 is the most abundant GPCR in rCFs

Using a RT-PCR GPCR array to identify nonchemosensory GPCRs expressed in cDNA generated from freshly isolated rCFs, we discovered that these cells express 190 GPCRs with a median dC(t) value of 20.9 and identified PAR1 as the most abundant receptor (Fig. 1 and Table 2; see Materials and Methods, RT-PCR GPCR Array). Among the most abundant receptors are the sphingosine-1-phosphate receptor 3, a number of frizzled homologue receptors, the P2Y2 receptor, and several orphan GPCRs. By contrast, certain receptors that have been the focus of previous studies with CFs, e.g., angiotensin, endothelin, and β2-adrenergic receptors, are expressed to a much lower extent than are PAR1 receptors (Table 3). Expression of select GPCRs was confirmed by RT-PCR with independent primer sets, and the rank order of expression closely matched results from the GPCR arrays (Supplemental Table S1). Limited prior data have indicated expression or a functional role for PAR1 in cardiac fibroblasts (14–16), but its actions in inflammation and endothelial barrier function have demonstrated that PAR1 can regulate early stages of wound repair (17–19). Since PAR1 expression is >4-fold higher than the next most abundant GPCR in rCFs, we undertook studies to define its functional role.

PAR1 activation is profibrotic in rCFs

PARs are a unique class of GPCRs in that they are activated by proteolytic cleavage of their N termini, which exposes a peptide ligand responsible for the binding and activation of the receptor (20). PAR1, activated physiologically by the protease thrombin, can also be activated by synthetic peptide ligands [e.g., thrombin receptor activator peptide 6 (TRAP6; SFLLRN)] that present the amino acid sequence of this cleavage site, thereby mimicking the activity of thrombin.

To define the functional effects of PAR1 activation, we treated rCFs with thrombin or TRAP6 for up to 24 h. Both thrombin and TRAP6 elevated the expression of profibrotic markers in rCFs (Fig. 2): PAI-1 (which regulates ECM turnover; ref. 21), the myofibroblast marker αSMA (1), Cyr61, CTGF, Nov family member 1 (CCN1; aka Cyr61; the CCN family of proteins are important for fibroblast activation, migration, and ECM synthesis; ref. 22), and the profibrotic cytokine transforming growth factor β1 (TGFβ1; ref. 23). Thrombin and TRAP6 maximally and respectively increased expression of PAI-1 by 8.4- and 10-fold, αSMA by 2.7- and 3.8-fold, CCN1 by 14.7- and 6-fold, and TGFβ1 by 2- and 1.6-fold. Thrombin and TRAP6 decreased the expression of the exchange protein activated by cAMP (Epac1), an antifibrotic gene (24), by 2.1- and 1.9-fold, respectively (Fig. 2E, K). We tested whether the increased mRNA levels of PAI-1 and αSMA promoted by PAR1 activation are reflected by changes in protein expression. Thrombin and TRAP6 increased protein levels of PAI-1 (9.1- and 14.3-fold, respectively, with effects peaking at 8 h) and αSMA (2.3- and 2.7-fold at 24 h), results akin to the changes in mRNA levels (Fig. 3).

Certain markers (e.g., CCN1) were transiently affected, while other effects [PAI-1, αSMA, TGFβ1, and the decrease in exchange protein activated by cAMP 1 (EPAC1)] were more persistent, implying that PAR1 activation produces a complex profibrotic response in rCFs (Fig. 2F, L). In addition, thrombin and TRAP6 have different kinetics in their effect on the expression of the fibrotic markers. The effects of thrombin were more persistent than those in response to the peptide, indicating that TRAP6 does not fully mimic thrombin during long-term treatment. Despite these differences in the response of rCFs to thrombin and TRAP6, the overall response is profibrotic and similar to that observed for other profibrotic stimuli, such as angiotensin II and TGFβ1 (22, 24–26).

PAR1 activation induces morphological changes and increases cellular collagen synthesis in rCFs

To confirm the profibrotic activity of PAR1 activation on rCFs, we assayed the effects of thrombin on cell morphology, the production of ECM, and cellular proliferation. The activation of fibroblasts and transformation to the myofibroblast phenotype produces morphological changes that we assessed using immunofluorescence microscopy (Fig. 4A–C). Treatment with thrombin (4.5 nM, 24 h) increased αSMA staining and induced reorganization of the actin network: the appearance of αSMA changed from diffuse cytosolic localization into defined actin stress fibers (Fig. 4B). Thrombin-treated cells also were generally larger and had a more rigid appearance. The myofibroblast phenotype induced on thrombin treatment of CFs was similar to that produced by the profibrotic agonist angiotensin II (1 μM, 24 h, Fig. 4C).

We assayed collagenase-sensitive [³H]proline incorporation as a measure of the effects of thrombin on ECM production and assessed cell proliferation by measuring [³H]thymidine incorporation. Incubation with thrombin for 24 h increased collagenase-sensitive [³H]proline incorporation by 60% (relative to untreated cells), a response similar to that produced by angiotensin II (Fig. 4D). Thrombin had a limited, statistically insignificant effect on [³H]thymidine incorporation, by contrast with the mitogenic effect of the control stimulus FBS (Fig. 4E). Thus, thrombin initiates a proliferation-independent increase in collagen synthesis in adult rCFs.

Profibrotic response of rCFs to thrombin is mediated by PAR1 receptors

We next tested if the responses to thrombin were mediated by PAR1. It is difficult to completely and selectively inhibit these receptors with small-molecule antagonists (at least in part as a consequence of the irreversible activation of PAR1 receptors by thrombin). We thus used a knockdown approach by transfection of a PAR1-specific siRNA. This approach reduced PAR1 mRNA expression by 90% compared with cells transfected with a nonspecific control siRNA (Fig. 5A). The knockdown of PAR1 had no effect on basal PAI-1 or αSMA expression but almost totally eliminated the thrombin-induced increase of these profibrotic markers relative to control siRNA-transfected cells (Fig. 5B, C). In addition, PAR1 knockdown prevented the thrombin-induced increase in collagenase-sensitive [³H]proline incorporation, which was observed in cells transfected with a nonspecific control siRNA (Fig. 5D). Thus, the profibrotic effects of thrombin on CFs are mediated by PAR1.