Diverse expression of selected cytokines and proteinases in synovial fluid obtained from osteoarthritic and healthy human knee joints

Osteoarthritis (OA) is characterized by slowly progressing destruction of the articular cartilage as well as an inflammatory reaction within the realms of the affected joint [1]. OA represents the most common joint disorder in the world and results in pain, deformity and gradual loss of joint function [1]-[3]. The origin includes mechanical injury, genetic factors and aging [4]. Disease progression is characterized by formation of osteophytes, subchondral bone cysts, increased thickness of the subchondral bone plate and loss of articular cartilage [5]. The exact etiology of OA, however, is not fully understood.

Synovial inflammation is commonly found in early stage OA and is characterized by joint swelling, redness, heat and pain [6],[7]. The synovium is infiltrated by an increased number of synovial lining cells and immune cells such as macrophages, mast cells, T- and B-lymphocytes and plasma cells. Activated synovial cells, immune cells and chondrocytes produce large quantities of matrix metalloproteinases (MMPs) -1, -3, -9, -13 and ADAMTS, a disintegrin and metalloproteinase with thrombospondin motifs [8]. These proteolytic enzymes eventually degrade extracellular matrix macromolecules and cause a loss of cartilage integrity and biomechanical strength [9]-[11].

Proinflammatory cytokines act as soluble mediators and favor catabolic activities in articular cartilage [12],[13]. Cytokine production is associated with certain immune cells while their specific origin in OA, however, is still unclear [14]. The proinflammatory cytokines IL-1β and TNF-α are found in elevated concentration in synovial fluid (SF) and joint tissue of OA patients. Both activate the JNK, p38 MAPK and NF-κB signaling pathways in chondrocytes [15],[16]. They are capable of suppressing proteoglycan, collagen II and aggrecan expression while stimulating MMP expression and the production of further catabolic cytokines such as IL-6, IL-8, MCP-1 and CCL-5 [17].

OA is no longer understood as a disease merely caused by wear and tear but rather as an inflammatory process [18]. Definition of baseline cytokine and proteinase levels and their mRNA concentration in SF of healthy and OA patients is crucial to evaluate the potential for diagnosing early OA using cytokines as biomarkers [19]. The objective of this study was to elucidate whether the cells in the SF are producing the detrimental proteins mentioned above. Thus, we aimed to quantify mRNA expression patterns of the main inflammatory mediators IL-1β and TNF-α and degradative enzymes MMP-1 and MMP-3 within synovial fluid of OA patients compared to healthy controls. Such information would further define the role of the mentioned cytokines and proteases within the complex cascades of incidental/progressive OA and could strengthen their potential as biomarkers and/or novel therapeutic targets.

Inflammation plays a pivotal role in the pathogenesis of OA [4]. However, the cellular origin of inflammatory mediators and catabolic enzymes remains an area of speculation [14]. The aim of this study was to quantify and compare inflammatory and degradative markers on mRNA levels in cells floating in the SF of OA patients and healthy controls. The selection of TNF-α, IL-1β, MMP-1 and MMP-3 represents well-established inflammatory mediators and catabolic proteinases in OA [15],[18]. The present study demonstrates that there is significantly more IL-1β, MMP-1 and MMP-3 mRNA transcribed in SF cells of OA patients compared to healthy controls. The mean age of OA patients was slightly older compared to the donor group and older age could lead to a different onset of inflammatory response. Furthermore, there was a difference in mean BMI comparing OA patients and donors. Higher BMI causes higher joint load and one could imagine that this might interfere with the inflammatory response of the investigated joints. However, the immense differences between the two groups concerning frequencies of detectable expression of IL-1β, MMP-1 and MMP-3 are not likely to be caused by the slight differences of patient age or BMI.

Increased levels of active IL-1β have already been found in SF of OA patients using ELISA [21]-[24]. Similar findings were made analyzing OA cartilage and synovial tissue using immunohistochemistry [25]-[27]. To our knowledge, this is the first study to analyze mRNA levels of IL-1β in SF cells of OA patients. Our data demonstrate a high inflammatory activity of SF cells via IL-1β. However, it is not clear to what extent these cells act synergistically with cells within the synovial tissue and cartilage. Furthermore, IL-1β’s mRNA still has to be translated and IL-1β’s inflammatory potential is influenced by the presence of IL-1β-converting enzyme, allocation of suitable receptors and other inflammatory processes. Our analysis of TNF-α mRNA concentration showed a wide range of results within the group of healthy controls and no significant difference to our OA group (P = 0.077). Although TNF-α is accepted as one of the two most important inflammatory cytokines in OA today, conflicting reports are found in the literature. Early studies showed highly inconsistent levels of TNF-α expression in SF and synovial tissue of OA patients [24],[28]-[30]. Furthermore, TNF-α levels in SF significantly correlate with clinical symptoms but not with radiographic findings of OA [31]. An answer to these contrary reports might be variations in number and susceptibility of chondrocyte TNF-α receptors, the presence of soluble TNF-α receptors and activity of TNF-α-converting enzymes [24]. Alternatively, TNF-α-expressing cells such as macrophages could mainly be localized in the synovial tissue [32].

IL-1β and TNF-α stimulate the production of cartilage-degrading enzymes [33],[34]. As such, MMPs are of particular interest because they are considered as the major degradative enzymes in OA [11]. Augmented MMP levels were found in SF of OA patients and they are expressed by lining cells, neutrophils, macrophages and chondrocytes in the synovial tissue [35]-[38]. Moreover, our data indicate that a significant amount of MMPs is present within the synovial fluid. Again, it is not clear whether these MMPs are produced by cells in the surrounding tissues or by cells within SF. First and foremost, our data indicate that cells in the SF are significant contributors. Thus, this study gives valuable information about potential treatment options including replacement of SF. Arthroscopic joint irrigation has been reported to cause significant release of pain in early studies. Livesely and colleagues showed more pain relief after arthroscopic lavage and physiotherapy compared to physiotherapy alone [39]. In contrast, Moseley and co-workers showed no difference between arthroscopic irrigation and sham operations of OA knee joints, nor regarding pain or function [40]. However, patients in the latter study were predominantly male and younger than our cohort leading to the thought that only certain patients may profit from arthroscopic irrigation. Removing inflamed SF at an early OA stage might inhibit the start of the self-escalating inflammatory cycle whereas joint lavage of late-stage OA might have no or only mild/temporary effects.

Early diagnosis of OA presents another clinical challenge and is crucial for disease prevention. Conventional diagnostic tools such as radiographic and magnetic resonance imaging (MRI) showed limited sensitivity or are time-consuming and expensive [41]. Analysis of cartilage matrix molecules in body fluids is still only used as research tool and gene expression profiles in peripheral blood leukocytes showed ambivalent cytokine levels [4],[42]. Alam and colleagues successfully showed in a canine animal model that biomarkers in SF such as tartrate-resistant acid phosphatase, MMP-2, and tissue inhibitor of MMP-2 can be used for diagnosis of early OA [43]. Another canine animal model by Garner and co-workers showed high sensitivity and specificity of IL-8 and monocyte chemoattractant protein-1 in SF to surgically-induced OA. Our data show that mRNA of IL-1β, MMP-1 and -3 in SF is a sensitive indicator for symptomatic OA [44]. Further research should evaluate whether these markers are also able to detect early or non-symptomatic OA and whether they can function as a prognostic tool.

Our findings show that SF of OA patients contains cells that actively promote inflammation via expression of proinflammatory mediators and joint-degrading enzymes. Although origin and type of cells still have to be investigated, this may offer an explanation as to why removal of OA SF can cause OA symptom relief. Furthermore, qPCR-based analysis of SF might offer a promising diagnostic tool and the data presented here lead to speculation towards potential pharmacologic OA intervention on a biological basis.