Mechanical injury and inflammatory cytokines affect cartilage integrity and tissue homeostasis : a mass spectrometric analysis of proteins with relevance to arthritis

Abstract

Osteoarthritis is characterized by synovial joint degeneration, and its cardinal pathological feature is degeneration and loss of the articular cartilage joint surface. While the aetiology of osteoarthritis is unknown, risk factors include gender, age, obesity, and prior joint injury. Joint injuries, including tears of the anterior cruciate ligament (ACL) and meniscus, increase the risk for the development of OA and involve both mechanical damage to cartilage, meniscus and synovial tissues, and tissue degradation associated with cytokine-induced inflammation. While the role of inflammatory cytokines in OA is still controversial, their role in rheumatoid arthritis is evidenced by the successful use of anti-TNF-ca and anti-IL-1 therapies to abrogate disease symptoms and progression. In vitro, both IL-13 and TNF-ca promote chondrocyte-mediated matrix degradation and inhibit cartilage matrix synthesis, while mechanical damage causes cell death, matrix damage, and decreased cell biosynthesis. Understanding the similarities and differences in cartilage responses to inflammatory cytokines and mechanical injury is important in understanding the catabolic-anabolic shifts that typify OA progression. Therefore, the objectives of this thesis were (1) to identify the role of TNF-a and IL-1 induced nitric oxide (NO) as a mediator of cartilage tissue damage;

(cont.) (2) to characterize and compare the regulation by IL- 1 P, TNF-cc, and mechanical injury of secreted factors, matrix degradation, and mechanisms of chondrocyte cell death using an SDS-PAGE-LC/MS/MS protein profiling approach; and (3) to further quantify the effects of IL-1 3, TNF-ca and injury using an isobaric isotope labeling (iTRAQ) based 2D-LC/MS/MS approach. Together these studies were designed to provide better understanding of matrix degradation, cell death, immune response, and evidence of cell-mediated repair processes. NO is produced by chondrocytes in response to inflammatory cytokines TNF-a, IL-1 3 and IL-17, and can mediate cellular and extracellular events through cGMP signaling, protein modifications (e.g., S-nitrosation or tyrosine nitration), altered transcript stability, and altered sugar and lipid chemistry. Cartilage was treated with IL-13 or TNF-ca left untreated in the presence or absence of the NO synthase inhibitor, L-N-methylarginine (L-NMA), and changes in gene expression and matrix breakdown were measured. We found that L-NMA treatment partially inhibited TNF-a-induced, aggrecanase-mediated aggrecan degradation as indicated by a decrease in sGAG loss to the medium and by an increase in the generation of aggrecanase-specific aggrecan fragments.

(cont.) No change was observed upon addition of L-NMA to IL- 1P treated explants, but addition of L-NMA to combined IL-1 3 and TNF-a treated explants increased sGAG loss, suggesting that the effects of NO may be contextual. We hypothesized that this might be due to differences in aggrecanase expression (ADAMTS4 vs. ADAMTS5) or post-translational modification, but no aggrecanase was consistently identified in the samples. No difference in MMP expression or activation was noted following addition of L-NMA, and no change in NO chemistry between IL-1 3 and TNF-a treatment was evident by nitrate and nitrite production. Gene expression analysis was conducted on a battery of 32 genes, including matrix proteins, inflammatory mediators, proteases, cytokines and growth factors, and housekeeping proteins. While IL-113 and TNF-a both increased the expression of proteases and inflammatory mediators, addition of L-NMA did not significantly affect expression of the genes tested. We concluded that the effects of TNF-a and IL-1 p-induced NO production may depend on differences in cellular responses to each of these cytokines and possibly to differences in signaling or aggrecanase expression.

(cont.) In the second study, newborn bovine calf cartilage explants were treated with 10 ng/ml IL-1 p, 100 ng/ml TNF-a, radially-unconfined injurious compression (strain: 50%; strain rate 1000/o/sec), or no treatment, and cultured for five days. Pooled medium was subjected to SDS-PAGE-LC/MS/MS, and data were analyzed by Spectrum Mill proteomics software, focusing on protein identification, differences between treatments and matrix protein proteolysis. Over 250 proteins were identified among the four protein groups including CD 109, platelet derived growth factor like protein, and scrapie responsive protein, which have not been previously identified in cartilage. IL-13 and TNF-a caused an increase in YKL39, YKL40, complement factor B, MMP-3, ECM-1, haptoglobin, serum amyloid A3, and clusterin. Injurious compression caused the release of intracellular proteins including GRP58, GRP78, alpha 4 actinin, pyruvate kinase, and vimentin, suggesting a loss of membrane integrity in a population of chondrocytes. Data on actin release within the first 24 hours suggested that this loss of membrane integrity occurred by mechanical cell disruption. Injurious compression also caused proteolysis of collagen type VI subunits, collagen type II, and COMP. Thrombospondin 1 fragments were seen in all treatment groups, and aggrecan proteolysis was predominant with cytokine treatment.

(cont.) Cartilage explants subjected to injurious compression released intracellular proteins and showed enhanced degradation of matrix proteins, while explants subjected to IL- 13 or TNF-ct released proteins involved in innate immunity and stress response. In the third study, cartilage explants were subjected to injurious compression, TNF-a (100 ng/ml) or IL-13 (10 ng/ml), or no treatment, cultured in equal volumes of medium, and the medium was collected, pooled and the proteins deglycosylated by treatment with chondroitinase ABC. The proteins were subjected to trypsinization, and the peptides were labeled with one of four iTRAQ labels each containing a unique signature ion. The labeled peptides were subjected to nano-2D-LC/MS/MS on a QStar, quadrupole time of flight instrument. The study was done in analytical replicate on a pooled sample of greater than 70 explants from a total of 6-12 different animals. Data were analyzed by ProQuant to obtain a ProGroup peptide report containing identified spectra, which were combined to achieve a peptide, and then a protein level output of mean ratios, standard deviations of those ratios, and significance based on either Wilcoxan sign rank or Student's t-test both corrected for multiple comparisons.

(cont.) Because of our interest in catabolic and anabolic shifts, a targeted data analysis approach was taken in addition to a systems level PCA and K-means clustering approach. By focusing on particular protein domains, we identified a decrease in the synthesis of most fibrillar collagen subunits (p<0.05), and an increase in the release of the aggrecan G2 and G3 domains with IL-13 and TNF-ta treatment (p<0.05). We also noted a significant increase in MMP-1, MMP-3, MMP-9, and MMP-13 in at least one condition and, in most cases, all conditions compared to the untreated sample. Increases in proteins involved in innate immunity and immune cell recruitment were noted with IL-1 [3 and TNF-a treatment, while an increase in intracellular protein release was seen most dramatically with mechanical compression injury. Since anabolic effects are often driven by the insulin-like growth factor family and the TGF-[3 superfamily, we specifically identified members of these pathways to understand which factors may mediate early repair processes.

(cont.) At the systems level, 2 principal components were sufficient to describe 97% of the covariance in the data. IL- 1 and TNF-a caused a similar response in proteins identified; in contrast, a 'Y'-shaped distribution was observed upon projection of proteins based on their response injury vs. cytokine treatment. K-means clustering revealed six main clusters to further characterize the biology of mechanical injury versus cytokine effects on cartilage.