Endoplasmic Reticulum Stress Regulates Rat Mandibular Cartilage Thinning under Compressive Mechanical Stress*

Background: The molecular mechanisms of mechanical stress-induced cartilage thinning remain largely unknown. Results: Endoplasmic reticulum stress (ERS) was activated in chondrocytes during mechanical stress loading. ERS inhibition suppressed the apoptosis and restored the proliferation and cartilage thinning. Conclusion: ERS regulates mechanical stress-induced cartilage thinning. Significance: Our data demonstrate a novel pathological role for ERS and provide new insight into the treatment of temporomandibular joint diseases. Compressive mechanical stress-induced cartilage thinning has been characterized as a key step in the progression of temporomandibular joint diseases, such as osteoarthritis. However, the regulatory mechanisms underlying this loss have not been thoroughly studied. Here, we used an established animal model for loading compressive mechanical stress to induce cartilage thinning in vivo. The mechanically stressed mandibular chondrocytes were then isolated to screen potential candidates using a proteomics approach. A total of 28 proteins were identified that were directly or indirectly associated with endoplasmic reticulum stress, including protein disulfide-isomerase, calreticulin, translationally controlled tumor protein, and peptidyl-prolyl cis/trans-isomerase protein. The altered expression of these candidates was validated at both the mRNA and protein levels. The induction of endoplasmic reticulum stress by mechanical stress loading was confirmed by the activation of endoplasmic reticulum stress markers, the elevation of the cytoplasmic Ca2+ level, and the expansion of endoplasmic reticulum membranes. More importantly, the use of a selective inhibitor to block endoplasmic reticulum stress in vivo reduced the apoptosis observed at the early stages of mechanical stress loading and inhibited the proliferation observed at the later stages of mechanical stress loading. Accordingly, the use of the inhibitor significantly restored cartilage thinning. Taken together, these results demonstrated that endoplasmic reticulum stress is significantly activated in mechanical stress-induced mandibular cartilage thinning and, more importantly, that endoplasmic reticulum stress inhibition alleviates this loss, suggesting a novel pharmaceutical strategy for the treatment of mechanical stress-induced temporomandibular joint diseases.

The temporomandibular joint (TMJ) 3 is a unique joint in the body; it is composed of stress-sensitive cartilage that is subject to extensive tissue remodeling (1). During orthodontic treatment, mechanical loading plays a pivotal role in the remodeling of mandibular cartilage and influences the growth of the mandible in the treatment of Class III malocclusion (mandibular prognathism) (2). Furthermore, various studies have shown that repeated increases in mechanical stress are closely associated with TMJ osteoarthritis. Cartilage thinning as a result of excessive mechanical stress has been recognized as a major factor in the development of TMJ osteoarthritis (3)(4)(5).
Mechanotransduction is the process of transmitting mechanical stimulation to cells. Previous studies have indicated that chondrocytes receive mechanical signals and integrate these signals with those from other signaling molecules, such as hormones and growth factors, thus determining cell fate. Several signaling pathways in chondrocytes are activated by mechanical stress (e.g. interleukin-4 (cytokine) signaling (6), intracellular calcium changes (7), and p38 MAPK phosphorylation (8)). However, the molecular mechanisms involved in cartilage thinning remain poorly understood.
Proteomic analysis provides a comprehensive view of cellular responses to various stimuli. This approach has been successfully utilized to examine chondrocyte differentiation and cartilage pathophysiology (9,10). Our previous study of the response of mandibular chondrocytes to mechanical stress in vitro demonstrated a transient but inhibitory effect on cell cycle progression, accompanied by cytoskeleton remodeling and MAPK pathway activation (11). However, the molecular mechanisms regulating mandibular cartilage responses to mechanical stress in vivo have not been thoroughly investigated, and the determinants mediating cartilage loss induced through mechanical stress remain largely unknown. * This work was supported by the National Natural Science Foundation of China and Nanjing Medical Science and Technique Development Foundation Contract Grants 81070807, 81200764, 81121062, and 50973046. □ S This article contains supplemental Table S1 and Figs. S1-S4. 1 To whom correspondence may be addressed. Tel.: 86-025-83620173; Fax: 86-025-83620173; E-mail: Lihuang76@nju.edu.cn. 2 To whom correspondence may be addressed. Tel.: 86-025-83620173; Fax: 86-025-83620173; E-mail: zchua@nju.edu.cn.
Here, we used an established rat model to load compressive mechanical stress on mandibular cartilage. The stressed chondrocytes were then isolated for a proteomic analysis. Proteins involved in endoplasmic reticulum stress (ERS), such as protein-disulfide isomerase (PDI), calreticulin (CRT), translationally controlled tumor protein (TCTP), and Pin-1 (peptidyl-prolyl cis/trans isomerase NIMA-interacting 1), were identified. As expected, the activation of ERS occurred during the mechanical stress loading. Furthermore, significant apoptosis was observed at 3 days post-stress, whereas the inhibition of cell cycle arrest and cartilage proliferation was observed later. Moreover, the inhibition of ERS reduced the observed apoptosis at the early stages of mechanical stress loading and restored the proliferation at the later stages. The results of this study provide new insight into the role of ERS in regulating cartilage thinning under conditions of mechanical stress and might facilitate the identification of new therapeutic targets for treating TMJ diseases.

Compressive Mechanical Stress Loading on Rat Mandibular
Cartilage-All animals were housed in an approved facility at Nanjing University. The animal use protocol followed institutional guidelines. Eight-week-old male Sprague-Dawley rats were used in this study. Compressive forces were loaded onto the TMJ as described previously (14) (see Fig. 1A). A rubber band was tied between the jig and the anchorage hooks to load 40 g of force on each side. Eight rats in each group wore the appliance for 3, 7, 14, or 21 days, with gender-and age-matched controls. None of the rats displayed signs of disability, and all animals received the same standardized diet throughout the procedure. To show the direction of mechanical stress loading, x-ray images were obtained for the control and experimental groups (30 kV, 3 mA, 1 min) (Fig. 1A).
Histological Observation and Histomorphometric Measurement-Rats were sacrificed by cervical dislocation under anesthesia. The TMJ was removed and fixed in 4% paraformaldehyde for 24 h. The specimens were decalcified in 15% EDTA solution for 8 weeks and embedded in paraffin after thorough rinsing. Sagittal sections 4 m in thickness were cut. The histological changes in mandibular cartilage were examined by hematoxylin-eosin (H&E) staining. Two sets of five stained sections from each experimental period were used for histomorphometric analysis. The entire mandibular cartilage area was divided into three zones (fibrous proliferation (FP), transition (T), and hypertrophic zones (H)) from the top to the bottom (Fig. 1B). This regionalization method has been extensively used in studies on mandibular condylar cartilage (12)(13)(14)(15). The thickness of each zone was measured with Image-Pro Plus software.
The 14-day experimental group and their age-matched controls were further examined using Alcian blue staining and type II collagen (1:100; rabbit anti-rat polyclonal antibody, KeyGen Biotech, Nanjing, China) immunohistochemical staining. The samples for the immunostaining of type II collagen were pretreated with hyaluronidase for 1 h.
Mandibular Condylar Chondrocyte Culture-Mandibular cartilage was obtained from Sprague-Dawley rats in the control and the experimental (14 days) groups and carefully dissected under a microscope. The dissected line is shown in black in Fig.  2A. The soft tissues of the mandibular condyle were separated using ophthalmic forceps and digested with 0.25% trypsin for 30 min. The mandibular condyle was subsequently minced and digested with 0.2% collagenase for 120 min. The primary mandibular chondrocytes were rinsed three times and prepared as a single-cell suspension in growth medium containing DMEM supplemented with 15% fetal calf serum and penicillin/streptomycin (Invitrogen). Cells were counted with a hemocytometer, and cell viability was verified with trypan blue solution. Subsequently, the cells were seeded overnight at high density (1 ϫ 10 6 /cm 2 ) in a humidified atmosphere at 37°C and 5% CO 2 . Primary mandibular chondrocytes were examined by inverted phase-contrast microscopy. The chondrocytes were evaluated by H&E staining and validated by type II collagen (1:200; rabbit anti-rat polyclonal antibody, KeyGen Biotech) immunohistochemistry following standard procedures. The chondrocytes were cultured 3 days to reach confluence.
Isolation of Total Proteins and Phosphoproteins-The cell pellets were lysed in a cell lysis solution containing 7 M urea, 2 M thiourea, 40 mM DTT, and 2% immobilized pH gradient buffer (pH 3-10 (Linear); GE Healthcare). The lysates were sonicated in short bursts to disrupt the nucleic acids and clarified by centrifugation at 40,000 ϫ g at 4°C for 60 min to obtain total proteins. The protein concentration in the supernatant was determined using the Bradford method. Phosphoproteins were isolated from mandibular chondrocytes using the phosphoprotein purification kit (Qiagen, Hilden, Germany) as recommended by the manufacturer (16). Briefly, 1 ϫ 10 7 cells were lysed in a cell lysis buffer containing detergent, nucleases, and protease inhibitors. Lysates were cleared by centrifugation at 40,000 ϫ g at 15°C for 90 min, and the supernatants were loaded onto phosphoprotein-specific columns for affinity chromatography. The eluted proteins were concentrated using Nanosep ultrafiltration columns (Pall, Dreeich, Germany).
Two-dimensional Gel Electrophoresis and MALDI-TOF MS-The protein lysates (120 g of total protein lysates and 80 g of the phosphoprotein fraction) were subjected to isoelectric focusing on a gel strip with a linear pH 3-10 gradient and resolved on an SDS-polyacrylamide gel. The separated proteins were visualized by diamine silver staining as described previously (17). After destaining with double-distilled water, the gels were scanned at 600 dpi resolution, and the images were processed with Adobe Photoshop software (Adobe Systems) and analyzed with Image Master Platinum TM software (GE Healthcare) according to the manufacturer's instructions. A greater than 2-fold change in the protein concentration was considered significant (p Ͻ 0.01; Mann-Whitney test). The differentially expressed proteins were extracted from the gel and digested as described previously (18).
MALDI samples were prepared according to a method described previously (19). Mass spectra were recorded with an Ultraflex II MALDI-TOF-TOF mass spectrometer (Bruker Daltonics GmbH, Bremen, Germany) with FlexControl TM version 3.0 software (Bruker Daltonics GmbH). The MALDI-TOF spectra were recorded in the positive ion reflector mode in a mass range of 700 -4000 Da, and the ion acceleration voltage was 25 kV. The acquired mass spectra were processed with FlexAnalysis TM version 3.0 software (Bruker Daltonics GmbH) with the following internal standards: peak detection algorithm, SNAP (sort, neaten, assign, and place); S/N threshold, 3; quality factor threshold, 50. The tryptic autodigestion ion picks (trypsin_[108 -115], MHϩ842.509, trypsin_[58 -77], MHϩ2211.104) were used as internal standards. Matrix and/or autoproteolytic trypsin fragments or known contaminant ions (keratins) were excluded. The resulting peptide masses were used to search NCBInr 20101105 (101,942 sequences) with Mascot (version 2.3.02) in automated mode, using the following criteria as search parameters: significant protein MOWSE score at p Ͻ 0.05, minimum mass accuracy of 120 ppm, trypsin as an enzyme, one missed cleavage site allowed, alkylation of cysteine by carbamidomethylation as a fixed modification, and oxidation of methionine as a variable modification. In addition, the Mascot score and expectation of the first non-homologous protein to the highest ranked hit were verified.
Isolation of Total RNA, RT-PCR, and Real-time Quantitative PCR-Total RNA was isolated from the control and stressed chondrocytes with the RNA simple Total RNA kit (Tiangen Biotech, Beijing, China), and reverse transcription was performed with the QuantScript RT kit (Tiangen Biotech) according to the manufacturer's instructions. cDNA was stored at Ϫ20°C. The following primer sequences were used to generate the products: CRT, forward (GAAATGGTGCTGGTCCTTC-TTCAC) and reverse (GTCACATGACACTCTCCACCAC-GA); TCTP, forward (ATTAGATATCCCCCCTCCCCCCG-CGCGCC) and reverse (CAGTGGATCCTTAACATTTCTC-CATCTCTAAGCCATC); PDI, forward (CTACGATGGCAA-ATTGAGCA) and reverse (CTTCCACCTCATTGGCTGTT); Pin-1, forward (TGCCCAGGCTTTTGTCAAAC) and reverse (CTCCAGTGCCAAGGTCTGAA); ␤-actin, forward (TCAG-GTCATCACTATCGGCAAT) and reverse (AAAGAAAG-GGTGTAAAACGCA). The PCR products were separated on a 1.5% agarose gel, and the expression of CRT, Pin-1, PDI, and TCTP was evaluated with Image-Pro Plus software (Media Cybernetics, Silver Spring, MD). For quantification, the mRNA expression of the above genes was normalized to that of the housekeeping gene ␤-actin.
Immunohistochemical Staining-Changes in the expression of PDI, CRT, phospho-Pin-1, TCTP, cyclin D1, PCNA, and Ki67 were observed by immunochemistry. The sections were rinsed with PBS. The antigens were retrieved by using an enzyme mixture containing trypsin and pepsin for 1 h at 37°C or by heating in a pressure cooker, rinsed with PBS buffer, treated with 3% H 2 O 2 in PBS, rinsed again with PBS, incubated with 10% normal rabbit or goat serum at room temperature for 10 min, and treated with one of nine antibodies, including CRT (1:100), TCTP (1:250), phospho-Pin-1 (1:50), PDI (1:100), cyclin D1 (1:100), rabbit polyclonal anti-Ki67 (1:100) (Abcam), and goat polyclonal anti-PCNA (1:100) (Zhongshan Inc.), in a moist chamber overnight at 4°C. After incubation with secondary IgG antibodies (HRP-conjugated) for 30 min at 37°C, the sections were rinsed and incubated for 30 min with SP complex at 37°C. The samples were stained with diaminobenzidine (SP immunochemistry kit, Boster Biological Technology, Ltd., Wuhan, China). The cell nuclei were counterstained with hematoxylin. The stained samples were embedded on microscope slides with neutral resins. A color difference was used to delineate positive and negative areas, which were measured with Image-Pro Plus software.
Flow Cytometry Analysis-Chondrocytes from the control and experimental groups (14 days) were fixed in 75% ice-cold ethanol and incubated with RNase A (200 mg/ml) in PBS at 37°C for 30 min. The fixed cells were subsequently washed twice with PBS, stained with propidium iodide in PBS, and analyzed by flow cytometry (BD Biosciences). The distribution of the cells in different cell cycle phases was indicated by the cellular DNA content profiles. A minimum of 10,000 events were counted for each sample. To detect apoptosis, annexin V and propidium iodide staining were performed using the annexin V-FITC/propidium iodide apoptosis kit (Biovision). The chondrocytes from the control and experimental groups were washed, trypsinized, and pelleted. The supernatant was removed, and the cells were resuspended in staining buffer con-taining the annexin V-FITC antibody and propidium iodide. The cells were analyzed by flow cytometry (FACSCalibur, BD Biosciences) with CellQuest software version 3.3 (BD Biosciences).
TUNEL Staining-Tissue sections (4 m in thickness) were prepared according to standard protocols for H&E staining. The apoptotic cells in each section (three sections per rat, eight rats in each group) were visualized using terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) according to the manufacturer's instructions (Boster Biological Technology, Ltd.).
Determination of Free Ca 2ϩ in Chondrocytes-Free Ca 2ϩ was measured in chondrocytes as described previously (20). Briefly, the chondrocytes were incubated with Fluo-AM (GenMed) according to the manufacturer's recommendations and visualized by immunofluorescence microscopy (Olympus). The mean fluorescence intensity was used to evaluate the relative level of cytoplasmic free Ca 2ϩ and calibrated to 1.0 in the control group.
ERS Inhibitor Injection in TMJ-A total of 40 8-week-old Sprague-Dawley rats were organized into four groups: control, mechanically stressed, control plus ERS inhibitor injection, and mechanically stressed plus ERS inhibitor injection. Salubrinal (Calbiochem) was dissolved in dimethyl sulfoxide and further diluted with saline. The anesthetized rats were given injections of salubrinal (15 l, 75 M solution) (21) into the articular capsules of the condyle (22) 30 min before compressive stress loading, and the injections were repeated every other day. After 14 days, the rats were sacrificed as described previously. Histological changes were observed using H&E staining, and the ultrastructure changes were determined by transmission electron microscopy (TEM). PCNA and Ki67 immunochemical staining were used to detect proliferation changes in the mandibular cartilage. The changes in apoptosis were observed using TUNEL staining after 3 days of mechanical stress loading.
Statistical Analysis-All experiments were repeated three times. The data were expressed as the means Ϯ S.D. and compared using a t test or analysis of variance, depending on whether the data were normally distributed. All statistical analysis was performed with SPSS version 11.0 software, and p Ͻ 0.05 was considered statistically significant.

Thinning of Mandibular Cartilage under Compressive Mechanical Stress Conditions and in Vitro Culture of Mechanically Stressed
Chondrocytes-To explore the mechanisms involved in cartilage thinning in vivo, we used an established animal model for mechanical stress loading on mandibular cartilage. The efficiency of the model was verified by x-ray analysis, which demonstrated that the mandible moved posteriorly and superiorly, similar to clinical observations (Fig. 1A, top). The H&E staining analysis showed that the chondrocytes in the mandibular cartilage appeared disordered and disarranged at 3 days compared with the controls, whereas the thickness of the transitional and hypertrophic zones appeared slightly reduced (Fig. 1, B and C, and supplemental Fig. S1A). At 7 and 14 days, the thickness of the cartilage and the numbers of chondrocytes were reduced significantly in all regions (Fig. 1, B and C, and supplemental Fig. S1A). In the 14-day experimental group, the thinning mandibular cartilage was almost one-third the thickness of the control cartilage, and the cell number was approximately one-half the cell number of the control group. At 21 days, the thickness of each zone appeared to have increased slightly (Fig. 1, B and C, and supplemental Fig. S1A). These results demonstrated that compressive mechanical stress reduces the number of chondrocytes and the thickness of mandibular cartilage and results in mandibular cartilage thinning.
Next, we chose the 14-day loading time point to study the changes in the mandibular cartilage in further detail. The cartilage in the control group appeared transparent, shiny, and elastic, whereas the cartilage in the experimental group appeared dry, dark, and lacking in elasticity, with reduced thickness and increased bone sclerosis (Fig. 2A). Alcian blue staining confirmed the changes in the thinning cartilage and revealed a reduction in the amount of extracellular matrix (Fig.  2, B and E). Furthermore, the strong collagen II immuno-staining was lost in the experimental group (Fig. 2, C and E). Expanded bone marrow space and trabecular changes were observed (Fig. 2D), indicating the erosion of the mandibular subchondral bone. In vitro, the weight of the mandibular cartilage was reduced from 0.026 Ϯ 0.005 g to 0.015 Ϯ 0.003 g (Fig. 2F).
To further explore the cellular responses of chondrocytes to mechanical stress, we isolated chondrocytes and performed H&E staining and collagen II immunostaining. The primary mandibular condylar chondrocytes in the control group were polygon-shaped and exhibited the typical characteristics of chondrocytes. In the experimental group, the numbers of chondrocytes were reduced to one-third the number of the control group (Fig. 2F). Although the size and shape of the chondro-FIGURE 2. Isolation and characterization of mechanically stressed chondrocytes. A, chondrocytes were isolated from the covered cartilage (above black lines) in control and 14-day samples. B, the cartilage was untreated or treated for 14 days, and the samples were subsequently stained with Alcian blue (Ab) (magnification ϫ100). C, the samples from C were subjected to collagen II (Col II) immunostaining. D, expansion of the bone marrow space and subchondral bone erosion were observed (magnification ϫ100). E, the quantitative analysis from B and C confirmed the observations in the extracellular matrix and the loss of collagen II immunostaining. *, p Ͻ 0.05; #, p Ͻ 0.01. F, the cell numbers and the survival rate were reduced in the 14-day group (#, p Ͻ 0.01). G, the primary chondrocyte cultures were evaluated by H&E staining (magnification ϫ100). H, the primary chondrocytes cultures were examined by inverted phase-contrast microscopy (magnification ϫ100). I, real-time quantitative PCR analysis of osteogenic, chondrogenic, and fibrogenic genes as indicated. Error bars, S.D.
To examine whether mechanical stress promotes the differentiation of chondrocytes, we determined the mRNA levels of different marker molecules, such as the chondrogenic genes type II collagen and type X collagen, the fibrogenic genes FSP1 and vimentin, and the osteogenic genes RUNX2 and OP (Fig.  2I). There was almost no significant change in the expression of FSP1 and vimentin, and the expression of type II and type X collagen decreased. These results suggest that the compressive force in our model had insignificant effects on fibroblasts, whereas reductions in chondrogenic cells were observed. Interestingly, the reduced expression of RUNX2 and increased OP mRNA suggested high metabolic activity and enhanced subchondral bone sclerosis.
Identification of Differentially Expressed Proteins by Proteomic Analysis-A proteomic analysis of the isolated chondrocytes was performed (Figs. 2B and 3A and supplemental Fig. S2) to identify the proteins responsive to compressive mechanical stimuli. The two-dimensional electrophoresis gels of whole proteins displayed 1867 Ϯ 61 protein spots and 1588 Ϯ 96 protein spots in each group (i.e. 3-12 and 14 -195 kDa, respectively), and the phosphoproteins gels displayed 472 Ϯ 32 and 454 Ϯ 48 proteins spots, respectively. The image analysis revealed that mechanical stress modified the expression of 72 spots in the protein gels and 36 spots in the phosphoprotein gels, of which 25 significantly changed proteins (p Ͻ 0.01) were identified by mass spectrometry (supplemental Table S1).
The bioinformatics analysis (data not shown) identified interesting proteins that are directly or indirectly related to ERS. For example, ER molecular chaperones, such as PDI, CRT, and reticulocalbin-1, were highly expressed during ERS. The ER-resident protein Pin-1 and the Ca 2ϩ -binding protein TCTP were down-regulated in response to ERS. Thus, these data suggest that ERS occurs in chondrocytes under compressive mechanical stress. To confirm the proteomics data, we examined the expression of PDI, CRT, TCTP, Pin-1, and phospho-Pin-1 (Fig. 3B). The expression of phospho-Pin-1 and TCTP was dramatically reduced (p Ͻ 0.01), whereas the expression of PDI and CRT was increased after mechanical stress treatment (p Ͻ 0.01). The results of the RT-PCR and Western blot analysis indicated similar alterations in the proteomic analysis of these proteins.
Time-dependent Changes in Differentially Expressed Proteins in Vivo-We next analyzed the histological sections of mandibular cartilage using in situ immunohistochemistry (Fig. 4, A and  B, and supplemental Fig. S3, A and C). PDI expression increased rapidly at 3 days, maintained a high level of expression at 14 days, and recovered by day 21. The expression of CRT was similar, but the response was later compared with PDI. CRT expression peaked at 14 days. The increased expression of PDI and CRT was significant in the transitional and hypertrophic zone. The expression of phospho-Pin-1 was reduced by half at 3 days, reached its lowest level at 7 days, stabilized at 14 days, and recovered at 21 days. The expression of phospho-Pin-1 was gradually reduced in the proliferative and transitional zones at 3 days but declined sharply in the hypertrophic zone at 7 days. A remarkable reduction in TCTP protein levels was observed in samples at 3 days post-treatment in the proliferative zone and remained at low levels at 21 days post-treatment. These results strongly suggest that ERS is induced by compressive mechanical stress in chondrocytes.
ERS Is Activated in Response to Mechanical Stress-To confirm the onset of ERS during mechanical stress loading, we detected the levels of established ERS-associated markers (23,24). Bip/GRP78, phospho-eIF2␣, and CHOP were up-regulated in the mechanical stress-treated samples compared with the controls. Accordingly, procaspase-3 levels were reduced, indicating that caspase-3 cleavage and activation occur in mechanically stressed chondrocytes (Fig. 5, A and B).
Increased cytoplasmic calcium distribution has been observed during ERS (24). Therefore, we determined the cytoplasmic Ca 2ϩ levels in mechanical stress-treated chondrocytes (Fig. 5, C and E). As expected, the cytoplasmic Ca 2ϩ levels were significantly elevated after mechanical stress treatment (p Ͻ 0.01). Previous studies have shown that ERS induces ER membrane expansion (25,26). TEM analysis revealed that the ER membranes were expanded under mechanical stress (Fig. 5D), indicating that ERS occurred in these processes. These results provide direct evidence that ERS is induced upon mechanical stress treatment.
Mechanical Stress-induced Cell Cycle Arrest, Proliferation Inhibition, and Apoptosis as a Function of Loading Time-To explore the mechanisms underlying chondrocyte loss and cartilage thinning, we examined the occurrence of apoptosis and proliferation during mechanical stress loading. We examined apoptosis with the TUNEL assay. As shown in Fig. 6, A and B, dramatic cell death was observed at 3 days post-treatment. Positive staining was predominantly observed in the transitional and hypertrophic zone but not in the proliferative zone. Cell death was largely reduced in the samples at 7 and 14 days posttreatment. The effects of mechanical stress on chondrocyte proliferation were further examined. Cell cycle analysis by flow cytometry indicated that more cells were observed in the G 0 /G 1 phase compared with the controls at 14 days post-treatment (p Ͻ 0.01) (Fig. 6, C and D), indicating inhibition of proliferation. Indeed, the expression of cyclin D1 was significantly reduced in isolated chondrocytes at 14 days post-treatment (Fig. 6E). Similarly, reduced cyclin D1 expression was observed in the cartilage during mechanical stress loading (Fig. 6, F and  G). Furthermore, PCNA expression was reduced to 20% compared with the control group at 3 days, and this low level of expression remained at 7 and 14 days post-treatment (Fig. 6, F  and G). These results indicated that the proliferation of mandibular cartilage was continuously inhibited during mechanical stress loading, whereas apoptosis occurred primarily during early loading. These changes provide an explanation for the observation of cartilage thinning after mechanical stress loading.
Inhibition of ERS Alleviates Mandibular Cartilage Thinning-ERS typically induces apoptosis and inhibits proliferation in many physiological and pathological processes (23,24); therefore, we proposed that the induced apoptosis and proliferation inhibition observed under mechanical stress might reflect ERS. To test this hypothesis, we utilized an ERS-selective inhibitor, salubrinal, to examine the effects of ERS inhibition on the cartilage homeostasis response to mechanical stress. The efficiency of salubrinal treatment was confirmed in primary chondrocytes by Western blot analysis, as described previously in an in vitro ERS study (20,27) (supplemental Fig. S4). In the present study, the histological analysis (Fig. 7, A and B) revealed that the thickness of the mandibular cartilage in the salubrinal/stressed group recovered, particularly in the hypertrophic zone, and the arrangement was more organized. The thinning mandibular cartilage was restored from one-third to two-thirds the thickness and the cell numbers of the control cartilage. However, reductions in the extracellular matrix and changes in the subchondral bone showed no significant recovery after salubrinal treatment (data not shown). TEM analysis revealed the resto- ration of expanded ER membranes under mechanical stress after the inhibition of ERS (Fig. 7C).
Moreover, the inhibition of ERS by treatment with salubrinal significantly increased the expression of proliferative proteins, such as PCNA and Ki67, in the 14-day post-stress treatment samples obtained from the experimental group. The expression of PCNA and Ki67 was reduced to one-third and increased or recovered to two-thirds that of the control cartilage after salubrinal injection (Fig. 7, D, E, and G). The early apoptosis observed after mechanical stress for 3 days was significantly suppressed after salubrinal injection (Fig. 7, F and G). These results demonstrate that the inhibition of ERS regulates cartilage thinning through the induction of apoptosis and inhibition of proliferation and suggest that ERS plays an important role in pathogenesis under compressive mechanical stimulation.

DISCUSSION
Chondrocyte proliferation and the thickness of the mandibular cartilage are modulated by mechanical stimuli. Further-more, many studies have indicated that cartilage thinning is important in the pathogenesis of TMJ diseases (14, 28 -31). Teramoto et al. (14) demonstrated a significant decrease in the proliferation of condylar cartilage and the amount of extracellular matrix under compressive force, and we also previously demonstrated reduced proliferation of mandibular chondrocytes in response to mechanical stress in vitro (11). The results of the present study are consistent with these studies. In addition, the present study revealed the pathological change characterized by prominent cartilage thinning, reduced chondrocyte numbers, low proliferation, extracellular matrix degradation, and subchondral bone erosion (32,33).
Mechanical force or physical stress has "nonspecific" effects on cells. Emerging studies have indicated that mechanical stress-induced specific cellular signaling, referred to as mechanotransduction, plays important roles in normal development and tissue dysfunction and degeneration. Although several cellular events, such as calcium oscillation and extracellular matrix and actin reorganization, have been identified during mechanical stress, the specific molecular mechanisms underlying these processes, particularly mechanical stress-induced cell death, remain largely unknown. In the current study, we identified several candidate proteins, including PDI, CRT, TCTP, and Pin-1, that are involved in ERS and exhibit down-regulated protein or phosphorylation levels during mechanical stress treatment. Further examination by transmission electron microscopy indicated that ER membrane expansion occurs during stress treatment. These data indicate that ERS accompanied by severe cell death is induced in rat mandibular cartilage in response to mechanical force in vivo. Importantly, the application of an ERS-specific inhibitor efficiently suppressed cell death. ERS-induced apoptosis plays a critical role in the progression of neurodegenerative diseases. The data indicate that ERS-induced cell death also plays an important role in the mandibular cartilage thinning induced by mechanical stress. Whether other apoptotic pathways are involved in mechanical stress-induced cell death remains unknown. Taken together, these data provide a novel mechanism for chondrocyte apoptosis induced by mechanical force and suggest a novel signaling target for mechanical stress.
These data also demonstrate that the isolated chondrocytes exhibited altered cell cycle progression during mechanical stress treatments. Indeed, previous studies have shown that eIF2␣ (translation elongation initiation factor 2␣) inhibits the translation of cyclin D1, which in turn arrests cells in the G 1 /S transition. Interestingly, the use of salubrinal, a specific inhibitor of eIF2␣, in the present study restored the levels of Ki67 and PCNA in chondrocytes after mechanical stress treatment. These data suggest that eIF2␣ is a critical effector for the arrested cell cycle progression induced in chondrocytes by mechanical stress, implicating this effector as a pharmaceutical target for the study of mechanical stress-induced mandibular cartilage thinning.
Consistent with previous studies, these data also identified other candidates involved in cytoskeleton reorganization and cell metabolism during mechanical stress. It has been shown that TCTP expression is reduced in response to ER stress. Furthermore, the down-regulation of Pin-1 expression has also been observed under ER stress conditions. The data obtained in the present study also indicate that levels of TCTP expression and Pin-1 phosphorylation are reduced in response to mechanical stimuli. These data suggest other targets for ERS in chondrocytes that merit further study. Taken together, our data demonstrate that ERS is a critical intracellular target in mechanically stressed chondrocytes and indicate a novel role of ERS in mandibular cartilage thinning induced by mechanical stimuli.