Different Secretory Activity of Articular and Subcutaneous Adipose Tissues from Rheumatoid Arthritis and Osteoarthritis Patients

Rheumatoid arthritis (RA) and osteoarthritis (OA) are common joint diseases affecting millions of people worldwide [1, 2]. Although both diseases are characterized by joint and systemic inflammation, their pathogenesis is different. Rheumatoid arthritis is a destructive autoimmune disorder triggered by combination of genetic and environmental factors, characterized by chronic inflammation, supported by cells and factors of innate and adaptive immunity, resulting in a clinical picture—synovitis, cartilage, and bone damage as well as high-grade systemic inflammation which affects many organs outside the joint and causes complications such as cardiovascular disease [1, 3, 4]. Osteoarthritis is a frequently occurring musculoskeletal disorder causing stiffness, swelling, instability, and pain of the joints [5]. Risk factors are numerous, including age, female gender, genetic predisposition, anatomical structure, joint injuries, obesity, or metabolic syndrome [2, 5, 6]. Osteoarthritis often develops due to mechanical overloading of the joints accompanied by an imbalance between tissue damage and repair [7]. It is characterized by cartilage degradation, subchondral bone sclerosis, osteophyte formation as well as some degree of synovitis that may be either primary or secondary symptom [8, 9]. The disease affects also the soft tissue structures (joint capsule, menisci, ligaments) and is characterized by low-grade systemic inflammation [2, 7, 10]. Despite the progress in the treatment, both diseases are still incurable and lead to joint destruction and disability [2, 11‐14].

Accumulating data suggest contribution of the white adipose tissue (WAT) to the pathogenesis and/or progression of different disease, including RA and OA [6, 15]. Adipose tissue, recognized now as the main body endocrine organ, contains various cell types (adipocytes, fibroblasts, macrophages, and lymphocytes) as well as components of the vascular and nervous systems [6, 15, 16]. These cells produce many biologically active substances (adipocytokines), including pro- and anti-inflammatory cytokines, classical adipokines, growth factors, complement components, and many others that act in a para-, endo-, and autocrine ways and control the metabolism and immune processes in antagonistic or synergistic manner [17‐19]. Fat depots are heterogeneous in respect of body locations, features and functional characteristics, and their secretory activity is modified by inflammatory environment [20]. There are data suggesting contribution of intraarticular fat pad (IPFP, or Hoffa’s fat pad) and abdominal subcutaneous fat tissue to joint and systemic inflammation, respectively [6, 15, 21, 22].

The aim of present work was to further support contribution of WAT to rheumatic disease pathology by investigating whether basal and pro-inflammatory cytokine-triggered secretory activities of subcutaneous (ScAT) and intraarticular (AAT) adipose tissues obtained from OA and RA patients reflect differences between these diseases in an intensity of systemic and local inflammation and contribute to these pathological processes.

Rheumatoid arthritis and osteoarthritis are characterized by local joint inflammation and systemic inflammation of high- and low-grade, respectively. Thus, disease-specific pro-inflammatory environment can affect both articular a well as subcutaneous adipose tissue located near the joint. Previous publications have indicated that WAT is a very active tissue-secreting adipocytokines that can participate in the development and course of inflammation not only inside the joint but also throughout the body in both RA and OA. These reports have shown that in most cases adipocytokine production is significantly higher in articular than in subcutaneous adipose tissues [21‐24]. In order to broaden and systematize knowledge about contribution of adipose tissue to these diseases pathology, we decided not only to compare the secretory activity of both AAT and ScAT and to assess their ability to react to the inflammatory stimulus, but also to compare tissues obtained from patients with RA and OA. We assumed that we would find a number of differences between RA and OA in the activity of AAT and ScAT, reflecting dissimilarities of these diseases background, course and intensity of inflammation. Such information would be also useful in pointing out adipose tissue as a possible therapeutic target in these rheumatic diseases.

Consistently with previous reports [21, 22, 24], we found, both in RA and OA, higher spontaneous production of some factors (MMP-3, IL-1Ra, TGFβ) in AAT than ScAT (Figs. 1 and 2). MMP-3 is an enzyme involved in homeostatic wound healing as well as in pathological processes inside affected joints—its role in the degradation of joint cartilage is well known [22, 25]. The higher level of MMP-3 in AAT in both studied diseases is consistent with local destructive role of this enzyme. Higher production of anti-inflammatory TGFβ and IL-1Ra (natural IL-1β antagonist) may be an attempt to self-regulate and reduce local inflammation. In addition, we noticed higher level of IL-6 secretion by AAT than by ScAT in OA, but not RA, patients (Fig. 1). Other authors reported up-regulation of IL-6 in AAT in both diseases [6, 22, 26]. Differences in RA clinical activity and treatment of patients may explain this discrepancy. Interleukin 6 exerts a variety of effects on both innate and adaptive immune cells, is involved in the regulation of inflammation and lipid metabolism. It is estimated that about 35% of circulating IL-6 is derived from adipose tissue [27]. Elevated levels of this cytokine, correlated with joint destruction, were detected in the serum, synovial fluid and fibroblast-like synoviocytes (FLS) of RA patients [1, 3]. Also in OA up-regulated serum concentration of IL-6 was reported to be associated with the progression of inflammation [6, 28]. Our results showing in OA patients larger amounts of this pro-inflammatory cytokine in AAT than ScAT seem to reflect the presence of much stronger inflammation inside the joint than outside it. In RA, however, similar amounts of IL-6 in both tissues seem to mirror high-grade inflammation not only in the joint, but also in neighboring tissues, and suggest equal contribution of AAT and ScAT to support it (Fig. 1, Table 2 and 3). Despite above differences in basal secretory activities of tested adipose tissues, present results revealed ScAT located nearby inflamed joints as an active tissue, producing significant amounts of adipocytokines (Table 2). Therefore, we believe that it can contribute to the development, maintenance, and regulation of inflammation in both RA and OA.

We also found that both adipose tissues derived from OA patients produced spontaneously larger amounts of TNF and CCL2/MCP-1 than analogous rheumatoid tissues (Tables 2 and 3). The role of TNF in inflammatory processes is well understood—it is critical for the development of inflammatory response and triggers production of other pro-inflammatory cytokines, including IL-1β and IL-6. Like IL-1β, TNF enhances the synthesis of proteolytic enzymes, downregulates their inhibitors, and reduces production of articular tissue matrix macromolecules [23, 29]. In adipose tissue, TNF is synthesized by adipocytes and infiltrating macrophages [27]. TNF is obviously a critical factor involved in the pathogenesis of RA [1]. Its important role has also been confirmed in OA [29]. CCL2/MCP-1 in turn is a chemokine which among other activities attracts macrophages and T cells into adipose tissue [22]. In addition, we stated that AAT from OA patients secreted significantly more MMP-3 than respective rheumatoid tissue (Table 3). The above differences between OA and RA patients in basal secretory activities of AAT and ScAT may be caused by inequalities in the treatment, as RA patients received DMARDs and/or glucocorticosteroids while OA patients were given NSAIDs only. It is possible that weaker anti-inflammatory therapy in OA was not sufficient enough to control activity of tested tissues in vivo. Higher levels of some adipocytokines (TNF, CCL2/MCP-1) in ScAT from OA than RA patients can also result from different body compositions, as obesity was more frequent in OA than RA group (Table 1). Obesity is known to be associated with low-grade inflammation of adipose tissue [21]. However, we failed to find significant correlation between secretory activity of tested adipose tissues and BMI of the patients. Therefore, dissimilarity of the patients’ treatment is a more possible explanation.

Interestingly, we noticed an important difference in the anti-inflammatory cytokines secretion between diseases as well. Both adipose tissues from OA patients secreted more IL-1Ra compared to RA group (Tables 2 and 3). This antagonist is a potent inhibitor of action of IL-1β—cytokine endowed with strong ability to activate adipose tissue [29]. Rheumatoid ScAT secreted in turn greater quantity of TGFβ (Table 2). By contrast to IL-1Ra, TGFβ is a pleiotropic regulator of immune response, which controls differentiation and function of innate and adaptive immune cells. It plays an important role in the development of several T cell lineages, including regulatory T cells while suppresses Th1 and Th2 cells [30]. Thus, it is likely that in OA IL-1Ra overproduced by AAT and ScAT exerts overall protective anti-inflammatory effect on adipose tissue while in RA considerable amounts of ScAT-originating TGFβ restrict rather high-grade systemic inflammation.

Finally, to mimic in vivo milieu we evaluated and compared reactivity of AAT and ScAT from both patients’ groups to pro-inflammatory IL-1β. It is known that IL-1β is produced in inflamed joints by synovial cells and triggers production of TNF and other adipocytokines [22]. As expected, in our in vitro experiments, IL-1β caused significant increase of almost all analyzed factors in both AAT and ScAT of OA and RA patients (Table 4). Interestingly, in OA reactivity of ScAT to pro-inflammatory IL-1β was significantly higher than reactivity of AAT as a greater increase of pro-inflammatory IL-6, MMP- 3, and CCL2/MCP-1 as well as anti-inflammatory IL-10 and TGFβ secretion was observed in this tissue (Figs. 3 and 4). By contrast, in RA the reactivity of AAT and ScAT to IL-1β did not differ (Figs. 3 and 4). These results suggest that in osteoarthritis ScAT, located in an inflammatory environment of lower intensity than AAT, maintains a greater ability to respond to an inflammatory agent. In RA, we failed to find such differences probably because both AAT and ScAT are exposed to high-grade, chronic inflammation. Importantly, comparison of ScAT and AAT from OA versus RA patients revealed that rheumatoid tissues are characterized by higher reactivity to pro-inflammatory IL-1β, at least in the case of CCL2/MCP-1 and MMP-3 secretion (Table 4), giving support to critical role of this stimulus in RA pathology. As mentioned above, for both ScAT and AAT of OA patients high-basal secretion of IL-1Ra is a characteristic (Tables 2 and 3). Therefore, this unique disease-specific and inflammatory-protective feature may explain lower reactivity of osteoarthritis adipose tissues to IL-1β and consequently lower-grade inflammation, comparing with RA.

It should be underlined that our study has several limitations. Firstly, due to ethical reasons, we were unable to compare activity of articular and subcutaneous adipose tissues obtained from RA and OA patients to respective tissues from healthy donors. Thus, we could only characterize tissues from diseases with different intensity of systemic and local inflammation. Secondly, we focused only on the secretion of selected cytokines known to be important in analyzed diseases development and activity. As was mentioned in the introduction, white adipose tissue secretes many active factors. According to available knowledge classical adipokines are also considered as the key players in inflammation, also in rheumatic diseases [31]. Leptin, adiponectin, resistin, visfatin, and others have been reported to be involved in RA and OA pathogenesis by their impact on cartilage [32], synovium [33], bone, adipose mesenchymal stem cells [34], and various immune cells [31, 35]. Thus, our work is only a step for understanding the complex process in the patients’ body.