Regulation of plasminogen activator inhibitor 1 expression in human osteoarthritic chondrocytes by fluid shear stress: Role of protein kinase Cα

Authors


Abstract

Objective

To test a fluid flow system for the investigation of the influence of shear stress on expression of plasminogen activator inhibitor 1 (PAI-1) in human osteoarthritic (OA) articular chondrocytes (from lesional and nonlesional sites) and human SW-1353 chondrocytes.

Methods

Human SW-1353 chondrocytes and OA and normal human articular chondrocytes were cultured on type II collagen–coated glass plates under static conditions or placed in a flow chamber to form a closed fluid-circulation system for exposure to different levels of shear stress (2–20 dyn/cm2). Real-time polymerase chain reaction was used to analyze PAI-1 gene expression, and protein kinase C (PKC) inhibitors and small interfering RNA were used to investigate the mechanism of shear stress–induced signal transduction in SW-1353 and OA (lesional and nonlesional) articular chondrocytes.

Results

There was a significant reduction in PAI-1 expression in OA chondrocytes obtained from lesional sites compared with those obtained from nonlesional sites. In SW-1353 chondrocytes subjected to 2 hours of shear flow, moderate shear stresses (5 and 10 dyn/cm2) generated significant PAI-1 expression, which was regulated through PKCα phosphorylation and Sp-1 activation. These levels of shear stress also increased PAI-1 expression in articular chondrocytes from nonlesional sites and from normal healthy cartilage through the activation of PKCα and Sp-1 signal transduction, but no effect of these levels of fluid shear stress was observed on OA chondrocytes from lesional sites.

Conclusion

OA chondrocytes from lesional sites and those from nonlesional sites of human cartilage have differential responses to shear stress with regard to PAI-1 gene expression, and therefore diverse functional consequences can be observed.

Osteoarthritis (OA) is the most common joint disease among older persons. The most common sites of clinical manifestations of degenerative cartilage in OA are the hips, knees, and spine. The symptoms of OA include pain, stiffness, and deformation of the knee joints (1). OA occurs more frequently in women and elderly individuals, but being overweight and engaging in a regimen of intense exercise are also factors that convey a high risk of OA. Excessive and repetitive mechanical stresses on the articular joint generate cartilage wear and may induce OA (2). Despite the widespread incidence of OA in the human population, its etiology is still largely unknown.

Cartilage is a supporting connective tissue composed of small numbers of chondrocytes and large amounts of extracellular matrix (ECM). The ECM synthesized by chondrocytes has a high water content and contains collagen, proteoglycan (PG), metalloproteinases, and other small molecules; it plays an essential role in cartilage structure and function (3). Many biochemical and genetic factors, as well as mechanical stress, can modify the connection between chondrocytes and the ECM and alter chondrocyte metabolism (4).

Plasmin is involved in various physiologic mechanisms, including thrombolysis, cell migration, metastasis, and arthritis formation. In addition, plasmin can activate metalloproteinases and plays an important role in modulating cartilage function (5). Urokinase plasminogen activator (uPA), tissue-type plasminogen activator (tPA), and their inhibitor, plasminogen activator inhibitor 1 (PAI-1), are present in the cartilage to modulate plasmin activation and the degradation of ECM. OA cartilage has been shown to display increased plasmin activity and elevated levels of uPA and tPA, as well as a decrease in PAI-1 expression (6). It has been demonstrated that the activity of uPA is associated with the degradation of cartilage, while the activity of PAI-1 is associated with the synthesis of cartilage during pathophysiologic processes (7); hence, it is important to understand the molecular mechanisms involved in the regulation of PAI-1 expression in human cartilage.

Protein kinase C (PKC) is an enzyme activated by inositol phospholipid hydrolysis and acts as a key enzyme for signal transduction in various physiologic processes (8–10). The PKC family is composed of several isoforms that are divided into 3 basic classes (conventional, novel, and atypical), according to the structure of the regulatory domains and the methods of activation (10). The role of PKC in articular chondrocytes is still unclear, especially when these cells are exposed to mechanical stress. Kimura et al (11) reported that PKCα can be detected in cultured bovine articular chondrocytes, that PG synthesis is stimulated in a concentration-dependent manner by a PKC activator and inhibited by a PKC inhibitor, and that PKCα-transfected chondrocytes produce elevated amounts of PG. In addition, PKCα appears in chondrocytes in the early stages of OA (12). It has been shown that PKC plays an essential role in the metabolism of PG by chondrocytes, and that mechanical stress exerted on chondrocytes or the cartilage matrix may modulate signal transduction of PKC (13).

There is evidence suggesting that abnormal mechanical loading may be detrimental to cartilage tissue (4, 14). Previous studies have shown that chondrocytes of the superficial and transitional zones are exposed to high and low fluid flow, respectively (15, 16), suggesting that mechanical shear stress may be a pathophysiologic process relevant in cartilage biology. Moreover, in cartilage tissue engineering, the development of chondrocytes/cartilage constructs is affected by different levels of shear stress, ranging from ∼1 dyn/cm2 to 23.2 dyn/cm2, revealing that hydrodynamic shear stress may alter intercellular signaling pathways in chondrocytes (17, 18). Despite the extensive number of studies on the effects of mechanical forces on chondrocytes, the detailed mechanisms that transduce the mechanical stimuli of shear stress to intracellular signaling, which ultimately leads to regulation of downstream gene expression, remain unclear. Thus, studying the mechanotransduction in chondrocytes in response to shear stress may help to elucidate the mechanisms underlying the focal nature of OA.

In the present study, we investigated the effects of fluid shear stress on the expression levels of PAI-1 messenger RNA (mRNA) by quantification of gene expression with real-time polymerase chain reaction (PCR). In order to elucidate the signaling pathways involved in the shear-induced regulation of PAI-1 expression in human chondrocytes, we determined the activation of PKCα and that of Sp-1 in response to shear stress and examined the effects of specific inhibitors or small interfering RNA (siRNA) specifically targeting these signaling proteins on shear-induced PAI-1 expression. Our findings provide a molecular basis for the mechanism by which fluid shear stress regulates signal transduction and PAI-1 expression in human articular chondrocytes, both from a chondrocytic cell line and from patients with knee OA.

PATIENTS AND METHODS

Primary culture of normal and OA human chondrocytes

Human OA articular chondrocytes were isolated from resected osteochondral specimens obtained during primary total knee arthroplasty from patients with advanced knee OA (n = 10; ages 65–80 years). The diagnosis in all patients fulfilled the American College of Rheumatology criteria for knee OA and corresponded to grade IV severity of knee OA according to the Kellgren/Lawrence classification system (19, 20). OA chondrocytes from sites of lesions were derived from residual pieces of degenerating cartilage (Outerbridge grades III–IV) (21) over the medial condyle of the femur, and OA chondrocytes from nonlesional sites were derived from cartilage pieces over the lateral condyle of the femur in the same patient. Normal human chondrocytes were isolated from healthy cartilage pieces of the distal femoral condyles obtained from patients undergoing above-the-knee amputation (n = 4; ages 67–79 years) (Figure 1A). This study was performed with the approval of the ethics committee of National Chiayi University and Chiayi Veterans Hospital, and all patients provided their informed consent.

Figure 1.

A, Schematic representation of the medial and lateral condyle of the femur, depicting the sites of extraction of normal human chondrocytes and human osteoarthritic (OA) chondrocytes from lesional and nonlesional cartilage. B and C, Plasminogen activator inhibitor 1 (PAI-1) mRNA expression in samples of normal human cartilage and lesional and nonlesional human OA cartilage, as detected by real-time polymerase chain reaction. Bars show the mean and SEM fold change in fluorescence intensity relative to OA chondrocytes from nonlesional sites, with results normalized to the levels of GAPDH (B) and 18S ribosomal RNA (C). ∗ = P < 0.01.

Primary chondrocyte cultures were generated from cartilage sections obtained from both the lesional and nonlesional areas of OA cartilage. The cartilage tissue samples were minced and washed in Dulbecco's modified Eagle's medium (DMEM) and subjected to 0.25% trypsin–EDTA digestion, followed by overnight digestion in 0.2% type II collagenase. The resulting cell suspension was filtered through 100-μm nylon meshes, washed repeatedly with phosphate buffered saline (PBS), and centrifuged at 250g for 5 minutes. After the cells had reached ∼70% confluence, they were seeded in monolayer, in DMEM supplemented with 10% fetal bovine serum (FBS), onto glass slides (Corning, Corning, NY) coated with type II collagen. To ensure that the cell phenotype was maintained, only first-passage cultured chondrocytes were used.

Reagents

All culture materials were purchased from Gibco (Grand Island, NY). Bovine type II collagen was purchased from BD Biosciences (San Diego, CA). Bisindolylmaleimide I (a pan-PKC inhibitor), calphostin C (an inhibitor that specifically targets the conventional and novel, but not atypical, PKC isoforms), and Gö 6976 (an inhibitor of conventional PKC isoforms) were purchased from Calbiochem (La Jolla, CA). Mithramycin A (an inhibitor of Sp-1 binding) was purchased from Biomol Research Laboratories (Plymouth Meeting, PA). Rabbit polyclonal antibodies against phospho-PKCα, PKCα, and Sp-1 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The specific siRNA for Sp-1 and those for the PKC isoforms, as well as the control siRNA (scrambled negative control containing random DNA sequences), were purchased from Invitrogen (Carlsbad, CA). All other chemicals of reagent grade were obtained from Sigma (St. Louis, MO).

Culture of SW-1353 chondrocytes

Human chondrosarcoma SW-1353 cells were obtained from American Type Culture Collection (Rockville, MD) and cultured in DMEM supplemented with 10% FBS. After the cells had reached confluence (1–2 × 105 cells/cm2), they were trypsinized and seeded onto glass slides (75 × 38 mm; Corning) precoated with type II collagen (50 μg/ml in 0.1N acetic acid). The medium was then changed to DMEM containing 2% FBS, and the cells were incubated for a further 24 hours prior to being used in the fluid flow experiment.

Fluid flow experiment

The glass slides were precoated with type II collagen at 37°C for 1 hour, and chondrocytes were then seeded onto glass slides for 24 hours. The glass slides with cultured chondrocytes were mounted in a parallel-plate flow chamber, as described in detail previously (22). The chamber was connected to a perfusion loop system and kept in a constant-temperature–controlled enclosure. The perfusate was maintained at pH 7.4 by continuous gassing with a humidified mixture of 5% CO2 in air. The fluid shear stress (τ) generated on the cells by flow was estimated to be 2–20 dyn/cm2, unless otherwise noted, using the formula τ = 6μQ/wh2, where μ is the dynamic viscosity of the perfusate, Q is the flow rate, and w and h are the width and channel height, respectively.

Real-time quantitative PCR

Total RNA preparation and the reverse transcription reaction were carried out as described previously (23). PCRs were performed using an ABI Prism 7900HT (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. Amplification of specific PCR products was detected using SYBR Green PCR Master Mix (Applied Biosystems). The designed primers in this study were as follows: for PAI-1, forward 5′-CAT-CCC-CCA-TCC-TAC-GTG-G-3′, reverse 5′-CCC-CAT-AGG-GTG-AGA-AAA-CCA-3′; for PKCα, forward 5′-ATT-CTA-TGC-GGC-AGA-GAT-TTC-C-3′, reverse 5′-TCC-TTC-TGA-ATC-CAA-CAT-GAC-G-3′; for Sp-1, forward 5′-GGT-GCC-TTT-TCA-CAG-GCT-C-3′, reverse 5′-CAT-TGG-GTG-ACT-CAA-TTC-TGC-T-3′; for GAPDH, forward 5′-GGG-GTC-ATT-GAT-GGC-AAC-AAT-A-3′, reverse 5′-ATG-GGG-AAG-GTG-AAG-GTC-G-3′; and for 18S ribosomal RNA (rRNA), forward 5′-CGG-CGA-CGA-CCC-ATT-CGA-AC-3′, reverse 5′-GAA-TCG-AAC-CCT-GAT-TCC-CCG-TC-3′. RNA samples were normalized to the levels of GAPDH and 18S rRNA. All primer pairs had at least 1 primer crossing an exon–exon boundary.

The real-time PCR was performed in triplicate in a total reaction volume of 25 μl containing 12.5 μl of SYBR Green PCR Master Mix, 300 nM forward and reverse primers, 11 μl of distilled H2O, and 1 μl of complementary DNA from each sample. Samples were heated for 10 minutes at 95°C and amplified for 40 cycles of 95°C for 15 seconds and 60°C for 60 seconds. Quantification was performed using the 2math image method (24), where the Ct value was defined as the threshold cycle of the PCR at which amplified product was detected. The ΔCt value was obtained by subtracting the Ct value of the housekeeping gene (GAPDH or 18S rRNA) from the Ct value of the gene of interest (PAI-1). The present study used the ΔCt value of controls as the calibrator. The fold change was calculated according to the formula 2math image, where ΔΔCt was the difference between the ΔCt value and the ΔCt calibrator value (which was assigned a value of 1 arbitrary unit).

Functional PAI-1 activity assay

The PAI-1 activity in the conditioned medium of the flow system was determined using a commercial enzyme-linked immunosorbent assay (ELISA) kit (Molecular Innovations, Southfield, MI), following the manufacturer's instructions (25). One unit of PAI-1 activity is defined as the amount of PAI-1 that inhibits 1 IU of human single-chain tPA, as calibrated against the international standard for tPA.

Western blot analysis

Chondrocytes were lysed with a buffer containing 1% Nonidet P40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), and a protease inhibitor mixture (phenylmethylsulfonyl fluoride, aprotinin, and sodium orthovanadate). The total cell lysate (50 μg of protein) was separated by SDS–polyacrylamide gel electrophoresis (12% running, 4% stacking) and analyzed using the designated antibodies and the Western-Light chemiluminescent detection system (Bio-Rad, Hercules, CA), as described previously (26).

Reporter gene construct, siRNA, transfection, and luciferase assays

The PAI-1 promoter construct (PAI-1-Luc) contains 800 bp of PAI-1 5′-flanking DNA linked to the firefly luciferase reporter gene of plasmid pGL4 (Promega, Madison, WI). DNA plasmids at a concentration of 1 mg/ml were transfected, using lipofectamine (Gibco), into SW-1353 cells when the cells had reached 70% confluence. The pSV-β-galactosidase plasmid was cotransfected to normalize the transfection efficiency. The cells were kept unstimulated as static controls or were subjected to shear stress experiments 48 hours after transfection. For siRNA transfection, SW-1353 cells, which had reached 70% confluence, were transfected with the designated siRNA, using the RNAiMAX transfection kit (Invitrogen).

Chromatin immunoprecipitation (ChIP) assay

After crosslinking cells with 1% formaldehyde, the cells were centrifuged and then resuspended in lysis buffer for 3 cycles of sonication, each for 15 seconds. Supernatants were recovered by centrifugation. Immunoprecipitation was performed overnight with specific antibodies against Sp-1 at 4°C with rotation. PCR was performed with primers that amplify the part of the human PAI-1 promoters that contains the Sp-1 binding sites; these primers were 5′-TCA-GCA-AGT-CCC-AGA-GAG-GG-3′ and 5′-GAT-GAA-CTC-ATG-TTC-CAG-CC-3′.

Sp-1 transcription factor ELISA

Nuclear extracts of cells were prepared as described previously (27). Equal amounts of nuclear extracts were used for quantitative measurements of Sp-1 activation, using commercially available ELISA kits (Panomics, Redwood City, CA) that measure Sp-1 DNA binding activity.

Cellular DNA content analysis

Cell viability was analyzed by flow cytometry. Briefly, cells were harvested in PBS containing 2 mM EDTA, washed once with PBS, and fixed for 30 minutes in cold ethanol (70%). Fixed cells were washed once in PBS and permeabilized with 0.2% Tween 20 and 1 mg/ml RNase A in PBS for 30 minutes. The cells were then washed once in PBS and stained with 50 μg/ml of propidium iodide (Roche, Basel, Switzerland). Stained cells were analyzed with a FACSCalibur system (BD Biosciences), and the data were analyzed using CellQuest software (BD Biosciences). At least 3 independent experiments were performed.

Statistical analysis

Results are expressed as the mean ± SEM. Statistical analysis was performed using an independent Student's t-test for comparisons of 2 groups of data, and using analysis of variance followed by Scheffe's test for multiple comparisons. P values less than 0.05 were considered significant.

RESULTS

Down-regulation of PAI-1 in human OA chondrocytes from lesional sites.

Expression of PAI-1 mRNA in human articular chondrocytes from the lesional and nonlesional sites of OA cartilage (n = 10) and from normal healthy cartilage (n = 4) (Figure 1A) were analyzed using real-time PCR. The internal reference genes GAPDH and 18S rRNA were used for normalization of the results of real-time PCR. Healthy cartilage showed wide variations in the expression of PAI-1 mRNA, reflecting individual differences. There was no significant difference between healthy cartilage and nonlesional OA cartilage. However, expression of the PAI-1 gene was significantly down-regulated in OA chondrocytes from sites of lesions in comparison with OA chondrocytes derived from nonlesional sites (P < 0.01) (Figures 1B and C), indicating that PAI-1 expression is modulated in OA chondrocytes. (The results in Figure 1B were normalized to GAPDH, while those in Figure 1C were normalized to 18S rRNA.)

Up-regulation of PAI-1 gene expression by moderate levels of shear stress.

To study PAI-1 mRNA expression under different shear stress conditions, human SW-1353 chondrocytes (used as a cell model) were exposed to different fluid shear stresses (2, 5, 10, and 20 dyn/cm2) for 0.5, 1, 2, 3, or 4 hours, and the changes in PAI-1 mRNA expression were analyzed by real-time PCR and normalized to the levels of GAPDH and 18S rRNA. At shear stresses of 5 or 10 dyn/cm2, the PAI-1 mRNA level began to increase after 1 hour of shearing and reached its highest level at 2 hours; thereafter, it gradually reduced to the level of the static control (Figures 2B and C). In contrast, a lower shear stress (2 dyn/cm2) or higher shear stress (20 dyn/cm2) had no significant effect on the PAI-1 mRNA levels (Figures 2A and D). (Data on exposure of the SW-1353 cells to different levels of shear stress for 2 hours with results normalized to 18S rRNA as the internal reference gene are available from the corresponding author upon request).

Figure 2.

Time course of expression of plasminogen activator inhibitor 1 (PAI-1) mRNA in human SW-1353 chondrocytes under shear stresses of 2 dyn/cm2(A), 5 dyn/cm2(B), 10 dyn/cm2(C), and 20 dyn/cm2(D). The chondrocytes were kept under static control (CL) conditions or were exposed to the different shear stress levels for up to 4 hours. RNA samples were isolated at the indicated time periods and subjected to real-time polymerase chain reaction. Bars show the mean and SEM fold change in fluorescence intensity relative to OA chondrocytes from nonlesional sites, with results normalized to the levels of GAPDH. ∗ = P < 0.05 versus static control.

A luciferase assay was conducted to confirm the effects of shear stress on PAI-1 gene transcription. The luciferase activity of SW-1353 cells exposed to shear stresses of 5 and 10 dyn/cm2, but not 2 and 20 dyn/cm2, for 2 hours was significantly higher than that under static control conditions (Figure 3A).

Figure 3.

Analysis of PAI-1 promoter activity by luciferase assay (A) and PAI-1 protein activity by enzyme-linked immunosorbent assay (B) under different levels of shear stress. SW-1353 chondrocytes were transfected with the PAI-1-Luc plasmid and exposed to 2–20 dyn/cm2 shear stress for 2 hours; results are expressed as the fold change relative to the static control, normalized to β-galactosidase activity (A). PAI-1 activity was assessed in conditioned medium from the flow system under the different levels of shear stress for 2 hours (B). Bars show the mean and SEM. ∗ = P < 0.05 versus static control; # = P < 0.05 versus 5 dyn/cm2 shear stress. See Figure 2 for definitions.

We further examined functional PAI-1 activity in SW-1353 cells under different shear stress conditions with the use of a functional PAI-1 activity assay (Figure 3B). PAI-1 activity in conditioned medium subjected to 5 and 10 dyn/cm2 of the flow system was higher than that under the static control conditions and at 2 and 20 dyn/cm2 shear stress. These data confirmed that 5 and 10 dyn/cm2 shear stress efficiently up-regulate PAI-1 transcription and protein activity. Notably, the effect of 10 dyn/cm2 was significantly higher than that of 5 dyn/cm2.

Mediation of the shear stress–induced up-regulation of PAI-1 expression by the PKCα pathway.

Activation of PKC is known to be an important participant in mechanically induced signaling cascades in chondrocytes (9). To determine the role of PKC isoforms in the shear-induced PAI-1 expression in SW-1353 chondrocytes, the cells were incubated with the specific inhibitor for pan-PKC (bisindolylmaleimide I, 20 μM), that for the conventional and novel PKC isoforms (calphostin C, 100 nM), and that for the conventional PKC isoforms (Gö 6976, 5 μM) (19) for 1 hour, and then exposed to shear stress of 10 dyn/cm2 for 2 hours in the presence of each inhibitor. Induction of PAI-1 by 10 dyn/cm2 of shear stress was significantly inhibited by bisindolylmaleimide I, calphostin C, and Gö 6976, as shown in Figure 4A with results normalized to GAPDH (results normalized to 18S rRNA are available from the corresponding author upon request), suggesting that the conventional PKC pathway is involved in shear-induced PAI-1 gene expression.

Figure 4.

Changes in PAI-1 expression and protein kinase Cα (PKCα) phosphorylation in SW-1353 chondrocytes. A, Effects of 1 hour of pretreatment of chondrocytes with DMSO or the PKC inhibitors bisindolylmaleimide I (Bis), calphostin C (Cal), or Gö 6976 (Go) on PAI-1 mRNA levels in SW-1353 chondrocytes before and after 2 hours of shear stress at 10 dyn/cm2. Controls were left untreated under static conditions (CL). B, Effects of 48 hours of transfection of siRNA targeting PKCα (si-α), PKCβ1 (si-β1), or PKCγ (si-γ) on the levels of PAI-1 mRNA in chondrocytes before and after 2 hours of shear stress at 10 dyn/cm2. Controls were cells without transfection under static conditions (CL) or cells transfected with control siRNA (si-CL). Bars in A and B show the mean and SEM fold change in fluorescence intensity relative to the static control, with results normalized to the levels of GAPDH. ∗ = P < 0.05 versus static control; # = P < 0.05 versus control under shear stress (DMSO in A, si-CL in B). C and D, Effects of duration of shear stress (2–60 minutes at 10 dyn/cm2) (C) and shear stress level (2–20 dyn/cm2 for 10 minutes) (D) on PKCα phosphorylation in SW-1353 cells, as assessed by Western blotting (top) and as the fold change in phosphorylation levels (bottom). Bars in C and D are the mean and SEM fold change in band density relative to static control, normalized to total protein levels. See Figure 2 for other definitions.

The involvement of the conventional PKC pathway in shear-induced PAI-1 expression was further confirmed by our experiments showing that PAI-1 expression was inhibited by transfecting the cells with PKC isoform–specific siRNA (100 μg/ml) prior to the application of shear stress at 10 dyn/cm2 for 2 hours. Transfection of SW-1353 chondrocytes with PKCα siRNA resulted in a significant inhibition of the shear-induced PAI-1 gene expression. Transfection of cells with PKCβ1 or PKCγ siRNA had a minor suppressive effect on the shear-induced PAI-1 expression (Figure 4B, showing results normalized to GAPDH; results normalized to 18S rRNA are available from the corresponding author upon request). Minor suppressive effects were also observed in cells transfected with PKCδ, PKCε, and PKCι siRNA (results not shown). (Data indicating the inhibitory effect of all of these siRNA on PKCα phosphorylation and the excellent gene-silencing efficiency and high cell viability after 48 hours of transfection are available from the corresponding author upon request.)

Induction of PKCα phosphorylation by shear stress.

To investigate the effect of shear stress on the mechanotransduction of PKCα, SW-1353 cells were exposed to 10 dyn/cm2 shear stress, or kept in static conditions as a control, for 0, 2, 5, 10, 30, and 60 minutes. The activation of PKCα was determined by Western blotting with specific anti–phospho-PKCα antibodies, with results expressed as the extent of phosphorylation. The phosphorylation of PKCα in SW-1353 chondrocytes increased rapidly (within 2 minutes) after exposure to 10 dyn/cm2 shear stress and reached a maximal level at 10 minutes; the levels of phosphorylation decreased thereafter but still remained higher than that in the static control at 60 minutes (Figure 4C). SW-1353 chondrocytes exposed to shear stress at 5 dyn/cm2 for 10 minutes also significantly increased the phosphorylation of PKCα, although to a lesser degree than that after exposure to 10 dyn/cm2 (Figure 4D). These findings show that PKCα phosphorylation is involved in the regulation of PAI-1 gene expression.

Mediation of Sp-1 activation by shear stress–induced PAI-1 expression.

Because the promoter region of the PAI-1 gene contains the Sp-1 binding domain, which is responsible for the modulation of gene expression (28), we tested whether Sp-1 activation is involved in the signal transduction pathway leading to shear-induced PAI-1 gene expression. SW-1353 chondrocytes were transfected with Sp-1 siRNA or incubated with the specific inhibitor for Sp-1 (mithramycin A, 100 nM) (29) for 1 hour, followed by application of shear stress of 10 dyn/cm2 for 2 hours. The shear stress–induced PAI-1 mRNA expression was significantly reduced by Sp-1 inhibition with mithramycin A and Sp-1 siRNA (Figure 5A, showing results normalized to GAPDH; results normalized to 18S rRNA are available from the corresponding author upon request), indicating that Sp-1 is involved in the regulation of PAI-1 gene expression.

Figure 5.

Roles of Sp-1 and protein kinase Cα (PKCα) in shear stress–induced PAI-1 mRNA expression and Sp-1 activation in SW-1353 chondrocytes. A, PAI-1 mRNA levels were determined in SW-1353 chondrocytes pretreated for 1 hour with vehicle control (DMSO) or the Sp-1 inhibitor mithramycin A (MMA) or transfected for 48 hours with control siRNA (si-CL) or siRNA for Sp-1 (si-Sp-1), and then exposed to 10 dyn/cm2 shear stress for 2 hours. Cells left unexposed under static conditions served as controls (CL). B, SW-1353 chondrocytes were left under static conditions (CL) or exposed to 10 dyn/cm2 shear stress for the amounts of time indicated. Sp-1 activation was then determined by chromatin immunoprecipitation (ChIP) assay (I) or by transcription factor enzyme-linked immunosorbent assay (II). Chromatin was immunoprecipitated with anti–Sp-1 antibody. One percent of the precipitated chromatin was assayed to verify equal loading (INPUT). C, SW-1353 chondrocytes were pretreated for 1 hour with DMSO or the PKC inhibitors bisindolylmaleimide I (Bis) or Gö 6976 (Go) or transfected for 48 hours with si-CL or specific siRNA for PKCα (si-α), PKCβ1 (si-β1), or PKCγ (si-γ), and then exposed to 10 dyn/cm2 shear stress for 30 minutes. Sp-1 activation was then determined by ChIP assay. D, SW-1353 chondrocytes were pretreated with DMSO, Bis, Go, or MMA for 1 hour or transfected with si-CL, si-α, si-β1, si-γ, or si-Sp-1, and then exposed to 10 dyn/cm2 shear stress for 2 hours. Cells were cotransfected with the PAI-1-Luc plasmid, and PAI-1 promoter activity was measured by luciferase assay. Bars show the mean and SEM fold change relative to static control. ∗ = P < 0.05 versus static control; # = P < 0.05 versus DMSO-treated cells under shear stress; & = P < 0.05 versus control siRNA transfection. See Figure 2 for other definitions.

We further evaluated shear stress–induced Sp-1 activation using a ChIP assay. SW-1353 chondrocytes exposed to 10 dyn/cm2 shear stress showed a time-dependent increase in Sp-1 binding activity on the PAI-1 promoter from 10 minutes to 30 minutes after exposure, and the effect lasted for at least 2 hours (Figure 5B, panel I). To further confirm these results, we performed quantitative analysis for Sp-1 binding activity using the Sp-1 transcription factor ELISA. Results of the ELISA also showed that exposure of SW-1353 chondrocytes to 10 dyn/cm2 shear stress increased Sp-1 DNA binding activity beginning at 10 minutes after exposure, and the activity remained elevated for at least 2 hours (Figure 5B, panel II).

Pretreatment of cells with PKC inhibitors or transfection of cells with PKCα siRNA significantly inhibited the increase in Sp-1 DNA binding activity, but there was no suppression of Sp-1 binding activity in cells transfected with PKCβ1 siRNA or PKCγ siRNA (Figure 5C). Application of 10 dyn/cm2 shear stress for 2 hours caused an increase in promoter activity in SW-1353 chondrocytes transfected with the PAI-1-Luc plasmid. Pretreatment of cells with bisindolylmaleimide I, Gö 6976, or mithramycin A or transfection of cells with PKCα siRNA and Sp-1 siRNA resulted in a marked inhibition of shear stress–induced PAI-1 promoter activity. However, transfection with PKCβ1, PKCγ, or control siRNA had little effect on the induction of PAI-1 promoter activity by shear stress (Figure 5D). These results provide additional evidence that the PKCα and Sp-1 pathways play an important role in the regulation of shear stress–induced PAI-1 expression in chondrocytes.

Differential regulation of PAI-1 expression by shear stress in human OA chondrocytes.

Because PAI-1 mRNA expression in human OA chondrocytes from the sites of lesions was significantly down-regulated in comparison with that in OA chondrocytes from nonlesional sites (Figure 1), we examined the effects of shear stress on the regulation of PAI-1 gene expression in human articular chondrocytes, using real-time PCR; the internal reference genes GAPDH and 18S rRNA were used to normalize the results. As shown in Figures 6A and B, exposure to 10 dyn/cm2 shear stress for 2 hours significantly increased the expression ratio of shear stress–induced PAI-1 mRNA to static control PAI-1 mRNA in normal chondrocytes and in OA chondrocytes derived from nonlesional sites, but the same level of shear stress had a minor effect on the PAI-1 mRNA expression ratio in OA chondrocytes from lesional sites. (The results in Figure 6A were normalized to GAPDH, while those in Figure 6B were normalized to 18S rRNA.) These data suggest that 10 dyn/cm2 shear stress can further increase the PAI-1 expression in normal and OA lesional chondrocytes. In contrast, chondrocytes from lesional cartilage, which already exhibits a reduction in PAI-1 expression, will lose the ability to respond to the same shear stress.

Figure 6.

Effects of 10 dyn/cm2 shear stress on PAI-1, PKCα, and Sp-1 mRNA expression, PKCα phosphorylation, and Sp-1 activation in samples of normal human cartilage and lesional and nonlesional osteoarthritic (OA) cartilage. A and B, Cells were kept under static control conditions (CL) or exposed to 10 dyn/cm2 shear stress (SS) for 2 hours. Results are expressed as the mean and SEM ratio of shear stress–induced mRNA expression to static control mRNA expression for each type of sample, normalized to GAPDH (A) and 18S rRNA (B). ∗ = P < 0.05. C, PKCα mRNA expression was assessed in lesional and nonlesional OA cartilage. Results are expressed as the mean and SEM fold induction, normalized to GAPDH (I). ∗ = P < 0.05 versus OA nonlesional chondrocytes. Phosphorylation of PKCα in human OA chondrocytes (nonlesional and lesional) was assessed by Western blotting of cells kept under static control conditions or exposed to 10 dyn/cm2 shear stress for 10 minutes (II). Results are the mean and SEM fold change in band densities relative to nonlesional static controls, normalized to total protein levels. ∗ = P < 0.05 versus nonlesional static controls. D, Sp-1 mRNA expression was detected in samples of lesional and nonlesional OA human cartilage by real-time polymerase chain reaction (I). Results are the mean and SEM fold induction, normalized to GAPDH. Sp-1 activation was assessed by chromatin immunoprecipitation (ChIP) assay (II). Human OA chondrocytes were kept under static control conditions or exposed to 10 dyn/cm2 shear stress for 30 minutes, and ChIP assay for Sp-1 was performed. Chromatin was immunoprecipitated with anti–Sp-1 antibody. One percent of the precipitated chromatin was assayed to verify equal loading (INPUT). See Figure 2 for other definitions.

The expression of PKCα mRNA was down-regulated in OA chondrocytes from the sites of lesions compared with OA chondrocytes derived from nonlesional sites (Figure 6C, panel I). Exposure of OA chondrocytes from nonlesional cartilage to 10 dyn/cm2 shear stress for 10 minutes increased the phosphorylation of PKCα, but this effect was not observed in OA chondrocytes derived from lesional cartilage (Figure 6C, panel II). Although the expression of Sp-1 mRNA appeared higher in OA lesional chondrocytes, the difference compared with OA nonlesional chondrocytes was not statistically significant (Figure 6D, panel I). OA chondrocytes from lesional cartilage and OA chondrocytes from nonlesional cartilage showed different Sp-1 DNA binding activities induced by shear stress (Figure 6D, panel II), again indicating that the PKCα/Sp-1 pathway plays an important role in the regulation of shear stress–induced PAI-1 expression in human OA chondrocytes.

DISCUSSION

Biochemical and genetic factors, as well as mechanical stress, contribute to the OA lesion in human cartilage by disrupting chondrocyte–matrix associations and altering metabolic responses in the chondrocyte (30). The major biochemical factors involved in the pathogenesis of OA include cytokines (e.g., interleukin-1β [IL-1β] and tumor necrosis factor α), chemokines (e.g., stromal cell–derived factor 1), metalloproteinases, tissue inhibitors of metalloproteinases, and the uPA/PAI-1 system (30). PAI-1 is the main physiologic inhibitor of uPA and tPA in vivo. Temporal changes in the expression and relative activity levels of the PA/PAI pair may influence cell function, either as a direct consequence of ECM-barrier proteolysis or by modulating cellular adhesive interactions with the ECM (31).

Cell type–specific synthesis and subcellular targeting of PAI-1 and uPA appear to play an important role in modulating the progression of OA (6). Multiple factors, e.g., IL-1β (32, 33), have been shown to regulate PAI-1 expression in chondrocytes, but the role of shear stress in regulating PAI-1 gene expression in chondrocytes is unclear. The present study assessing the SW-1353 human chondrocytic cell line demonstrated that shear stress regulates PAI-1 expression at the transcriptional level. Moreover, analysis of human PAI-1 promoter activity revealed that Sp-1 functions as the cis element for shear stress responsiveness via PKCα phosphorylation.

Chondrocytes show different responses to distinct levels of mechanical stress. Previous studies indicated that 5–10 dyn/cm2 shear stress has anticatabolic effects on articular cartilage, and that shear stresses higher than 16 dyn/cm2 can induce chondrocyte apoptosis, inflammation, and cartilage degradation (34–37). In our study, we assessed SW-1353 chondrocytes in a fluid flow chamber to investigate the effect of shear stress on the signaling pathway leading to PAI-1 gene expression. In this system, shear stresses of 5 dyn/cm2 and 10 dyn/cm2 were found to activate the PKCα signal transduction pathway, followed by an increase in PAI-1 promoter activity through regulation of Sp-1 DNA binding, and these events led to increased PAI-1 expression on chondrocytes. Healy et al (37) also showed that cyclooxygenase 2 (COX-2) expression in human chondrocytes is detected only at levels above the threshold shear stress level of 5 dyn/cm2, and a further increase in COX-2 levels occurs after exposure to 10 dyn/cm2. According to these findings, shear stress applied at a level of ∼10 dyn/cm2 effectively decreases the catabolic effects in chondrocytes. In contrast, abnormally high mechanical loading, such as 20 dyn/cm2 shear stress, may induce COX-2–mediated inflammation (37) and reduce PAI-1–mediated inhibition of uPA, tPA, and cartilage degradation.

The present study characterized a novel mechanism in which moderate levels of shear stress play an important role in the regulation of PAI-1 expression in normal human chondrocytes, but this regulatory response is deficient in human OA chondrocytes at sites of lesions. Thus, in normal chondrocytes and in OA chondrocytes from nonlesional sites, shear stress at 10 dyn/cm2 induces the up-regulation of PAI-1 expression via activation of the PKCα and Sp-1 signaling pathways. In contrast, human OA articular chondrocytes from lesional sites not only have significantly lower PAI-1 expression under static conditions (Figure 1) but also do not possess the ability to respond to moderate shear stresses to increase PAI-1 expression (Figure 6A).

The abnormal response of lesion-derived OA chondrocytes to excessive loading is a likely contributor to the dysregulation of chondrocyte function, favoring disequilibrium between the catabolic and anabolic activities of the chondrocyte in remodeling the cartilage ECM by the production of metalloproteinases and aggrecanase (38). The progressive degradation of the cartilage matrix that occurs in OA indicates that there is a local imbalance in the proteinase–inhibitor content (6). Shlopov et al (30) found that in the joint of the same OA patient, the levels of matrix metalloproteinase 13 (MMP-13) and MMP-1 were higher in chondrocytes from lesional sites than in chondrocytes from nonlesional sites, and their results suggested that aberrant mechanical forces led to elevated levels of MMPs to enhance degradation.

Although previous studies have shown that OA chondrocytes from lesional articular cartilage and those from nonlesional articular cartilage have different levels of gene expression under static conditions (19, 39), there have been no reports on their differential gene expression in response to external stimuli (e.g., shear stress). It is known that vascular endothelial cells contain mechanoreceptors on the cell surface, and several studies have shown that shear stress can alter the gene expression in endothelial cells through the activation of intracellular signal transduction pathways (40, 41). In primary culture human chondrocytes, we observed that the influence of moderate shear stresses on healthy and OA lesional chondrocytes was similar to the results in SW-1353 chondrocytes. In contrast, in OA lesional chondrocytes, such moderate shear stresses failed to activate PKCα and had no effect on the Sp-1 DNA binding activity; thus, this signal transduction pathway is suppressed in OA lesional chondrocytes. These findings serve to explain the differential regulation of PAI-1 expression on OA chondrocytes from lesional sites compared with those from nonlesional sites.

To respond to mechanical environments such as shear stress, chondrocytes may have putative mechanosensors on the cell surface, and different levels of shear stress may activate different mechanosensors and/or different signaling pathways. According to our findings, the process of mechanotransduction in response to moderate shear stresses may be malfunctional in chondrocytes obtained from the OA lesion.

It has been reported that shear stress can activate MAPK pathways to regulate early and late inflammatory responses in OA chondrocytes (12), but there is no report on the activation of PKC by shear stress in chondrocytes. PKCα plays an essential role in chondrocyte dedifferentiation and redifferentiation (42), and there are several reports on the possible linkage between PKC isoforms and the pathogenesis of OA. Hamanishi et al (43) found that the activation of PKC can inhibit the pathogenesis of OA in animal studies. Thus, the regulation of PKC in chondrocytes by shear stress could have physiologic and pathophysiologic significance. Our findings of the decrease in PKCα expression and the lack of a shear stress–induced phosphorylation response of PKCα in OA chondrocytes from lesional articular cartilage suggest that these molecular derangements may lead to dysfunction of the chondrocytes at lesional sites. The PAI-1 promoter has different binding sites for transcription factors such as Sp-1, Ets-1, and activator protein 1 (28). In the present study, we used a transcription factor ELISA and a ChIP assay to demonstrate that the regulation of PAI-1 gene expression on chondrocytes is mediated by the activation of Sp-1 and by an increase in Sp-1 DNA binding activity following PKCα phosphorylation. Nonlesional chondrocytes from OA articular cartilage can activate PKCα and increase Sp-1 DNA binding activity in response to shear stress, but such shear stress–induced effects are absent in OA chondrocytes from lesional cartilage.

Cartilage-specific matrix components (type II collagen) play a major role in the regulation of chondrocyte differentiation and the expression of chondrocyte phenotype (44). When grown on type II collagen, the chondrocytes maintain their round phenotype, produce type II collagen and fibronectin, and express β1 integrins on their cell membrane. Results from that previous study also indicated that the influence of type II collagen on cellular behavior depends on the integrins participating in a chondrocyte–type II collagen interaction (44). In addition, OA chondrocytes show a strong activation of synthetic activity in the increased expression of type II collagen (45).

Although the literature contains extensive reports on the effects of fluid shear stress on chondrocytes, the detailed mechanisms that transduce mechanical stimuli to intracellular signals to regulate the downstream gene expression remain unclear. Integrins, as the main receptors that connect the cytoskeleton and the ECM, have been shown to play important roles in transmitting mechanical stresses into chemical signals in a wide variety of cells seeded on the ECM (46). Recent studies indicate that shear stress–induced signal transduction and gene expression are dependent on the specific integrin and its cognate ECM protein to which the cells are adhered (47, 48). Therefore, elucidation of the role of integrins in these events requires further studies on OA lesional, OA nonlesional, and normal articular chondrocytes with the use of different ECM proteins and by manipulation of the integrin activities.

In summary, our present study revealed that, in comparison with nonlesional cartilage from OA patients and normal healthy cartilage, lesional cartilage from OA patients exhibits a decrease in PAI-1 mRNA expression and a loss of induction of PAI expression by shear stress through PKCα phosphorylation, Sp-1 activation, and Sp-1 DNA binding. Our findings provide a molecular basis for understanding the mechanisms contributing to the function of chondrocytes in a healthy state and to the dysfunction of chondrocytes in the progression of a disease such as OA.

AUTHOR CONTRIBUTIONS

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. Cheng-Nan Chen 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. Chien, C-N Chen.

Acquisition of data. Yeh, Chang, Chiang, Tsai, L-M Chen, Wu.

Analysis and interpretation of data. Yeh, Chang, Chiang, Tsai, L-M Chen, Wu.

Ancillary