Background
RA is a chronic autoimmune and inflammatory disease affecting synovial tissue in multiple joints. Pro-inflammatory cytokines are directly implicated in the pathogenesis of RA promoting synovial proliferation, hyperplasia and angiogenesis. This perpetuates inflammation, stimulates pannus formation and damage of bone and cartilage [
1,
2]. The proinflammatory cytokines interleukin 1 (IL1), tumour necrosis factor alpha (TNFalpha), and interleukin 6 (IL6), in particular, have been identified and intra-articular as well as systemic levels of these cytokines may to some extent reflect disease activity [
3-
6]. Treatment with anti TNF alpha inhibitors has been very effective as a biological therapy [
7]. Unfortunately, 40% of RA patients who receive a TNF inhibitor fail to achieve an adequate response [
8]. Other biological therapies including Tocilizumab [
9,
10], rituximab [
11] and abatacept [
12] have been successful in the treatment of RA, but the variable clinical response to these targeted therapies strongly highlights the heterogeneity of inflammatory process in RA. Another cytokine, IL-15 has also been related to the inflammatory process in RA. IL-15 is a pro-inflammatory cytokine which acts at the innate and acquired immune response [
13]. This cytokine recruits circulating memory T cells in the synovial membrane and may up regulate other pro-inflammatory cytokines [
14,
15]. These events include induction of TNF production through activation of synovial T cell and by macrophages via a cell contact dependent mechanism [
16], also the activation of Th17 lymphocytes driving IL17 production [
17]. Moreover, soluble IL-15 appears to be an important contributor to osteoclastogenesis promoting bone erosion [
18]. IL-15 mediates its biological effects by binding to a high affinity heterotrimeric receptor complex which is comprised of IL-15Ralpha, IL-2/15Rbeta and IL-Rgamma. IL-15Ralpha is a unique high affinity private subunit for IL-15. IL-15Ralpha may be secreted as a functional soluble molecule (s-IL-15Ralpha) and could behave as antagonist or agonist by forming a complex with IL-15 [
19-
21], also, it has been described that soluble IL-15Ralpha could bind to membrane bound IL-15 to induce a reverse signaling mechanism up-regulating the production of IL-6, IL-8, and TNFalpha by human monocytes and in the cancer cell line PC-3 [
22,
23]. These antecedents highlight the importance of this receptor in IL-15 biology and trigger previous studies of IL-15Ralpha in synovial fluids, where we have detected increasing levels of this receptor in synovial fluids from RA regards to OA [
24]. The aim of this work is to compare the expression profile of IL-15Ralpha, its ligand IL-15 and two validated cytokines targets in RA, TNF alpha and IL-6, in SF from RA regards to Osteoarthritis (OA) patients. In addition, we will assess inter-individual expression of these cytokines and will analyze how this cytokines are correlated.
Discussion
Due to RA is a syndrome comprising different pathogenic subsets, studying cytokine expression profile in inflammatory SF provides a unique opportunity to examine the microenvironment within the inflamed joint. In our current work we assessed expression of IL-15Ralpha, IL-15, IL-6 and TNFalpha in RA compared to OA, a rheumatic non autoimmune disease. Besides, we examined how expression of these cytokines inters individual’s patients.
In correspondence to previous studies, we found high varied levels of TNFalpha and IL-6 in patients with RA which were significant compared to the OA group [
27-
29]. Presence of soluble IL-15Ralpha has been less studied in RA, but it was previously detected as soluble receptor in serum from mice and humans [
30]; recently we detected soluble IL-15Ralpha in SF from RA and OA patients, which was the highest in RA. In current work, we confirmed these results with a higher number of patients. It is important to highlight that the ELISA performed to detect IL-15Ralpha used as a capture a previously described peptide which specifically binds to IL-15Ralpha and displace IL-15/IL-15Ralpha binding in a dose dependent manner [
25]. Therefore, we considered that detected IL-15Ralpha could not form complexes with endogenous IL-15.
Taking into account presence of IL-15Ralpha in synovial fluid, we were interested in to know how IL-15Ralpha is correlated to its ligand IL-15 and TNFalpha, an important proinflammatory cytokine validated as a therapeutic target in RA. We considered previous reports showing that IL-15 induces TNFalpha production by synovial cells. Also we measured IL-6 based on our previous report where we found correlation between IL-15Ralpha and L-6.
Regarding IL-15, we did not find significant differences for soluble IL-15 comparing RA and OA groups. Interestingly, we found high levels of IL-15 in 6 patients within OA group which corresponds to patients with early diagnosis of disease. Biological onset of OA is not clearly understood but evidence suggests that progression of cartilage degradation is mediated largely by pro-inflammatory cytokines [
31,
32]. Recent data suggest that innate immunity may be a driver of the OA process, thus, increased levels of the IL-15 protein are found in the synovial fluid of early knee OA patients when compared to end-stage OA [
33]. More recently it was reported that serum IL-15 levels are associated with severity of pain in patients with knee OA; this suggests that IL-15 might play a role in the pathogenesis of early OA [
34]. Disease duration was not considered as a variable in our study at the moment of including patients. Actually, in our study 60% of patients (18/29) with OA were 24 or fewer months from diagnosis which could explain the fact that we did not find significant differences in IL-15 between RA and OA groups, conversely as previously reported elsewhere [
35,
36].
Additionally, we studied the inter-individual cytokine expression in RA patients. A pronounced inter-individual diversity of expression patterns for these 4 cytokines was observed in SF. This finding corresponds to previously reported results of heterogeneity obtained when IL-6, TNF alpha, IL-1alpha and IL-1beta which were determined by immunohistochemistry in synovial tissue of RA patients [
37]. In current work, we identified 4 clusters of SFs corresponding to different patterns of concentration of these cytokines. Cluster 1 groups 7 patients with low levels of cytokines concentrations. Cluster 2 group patients with at least 3 increased cytokines, high levels of TNFalpha and IL-6 were observed in all patients included in this cluster. Additionally IL-15 or IL-15Ralpha was elevated, suggesting a high inflammation level. Cluster 3 includes patients negative to TNFalpha with high levels of IL-6 which could have high or mid levels of IL-15Ralpha and IL-15. Cluster 4 include patients with high or mid levels of IL-15 and IL-15Ralpha which are negative to IL-6 and TNF alpha, suggesting that the IL-15 system could play an important role in the inflammatory process in these patients, independent of IL-6 or TNFalpha.
Indeed, we found a correlation between IL-15Ralpha and DAS28 suggesting a relation between IL-15Ralpha level and disease activity. Other finding was that 11 of 15 patients positive RF are included in clusters 3 and 4. It is in agreement with a previous report of association between IL-15 and RF. So far, it has not been reported any association between IL-15Ralpha and RF.
Regarding disease duration, we could stratify patients into 2 groups being 1st group (n = 9) those with early disease (we consider early disease patients with <2 years from time of diagnosis); the 2nd group (n = 13) comprises those patients > 10 years from time of diagnosis. Both groups were statistically analyzed and no differences were found. There was none association between disease duration and IL-15Ralpha. Interestingly, we found that in C3B, one of the two subgroups of cluster 3 and including patients with high level of IL-15Ralpha, 83% of them had late disease.
Heterogeneity observed in the expression pattern of the studied cytokines should be taken into account when cytokines are used as targets in a therapeutic approach. In our current study, Cluster 3 and Cluster 4, which include 15 of 30 patients with signs of inflammation, are negative to TNFalpha. This could explain the fact that even when inhibiting TNFα in patients with active RA which results in a rapid and sustained improvement of their signs and symptoms of disease, it has been reported that more than 40% of them do not respond to this treatment [
38]. In patients treated with Tocilizumab a humanized monoclonal antibody against IL-6 receptor, also has been reported about 40% of inadequate response [
39]. Previous studies have shown a positive correlation between TNFalpha levels in serum and response to therapies inhibiting TNFalpha, although there are reports with the opposite result [
40-
42]. Frequently, cytokine levels in serum do not correspond with cytokine levels present in synovial fluid.
Cluster analysis has also shown a positive correlation between IL-15Ralpha and IL-6. Additionally we performed correlation analysis between the four studied cytokines. Correlation between IL-15Ralpha and IL-6 was confirmed and interestingly, we found also correlation between IL-15Ralpha and IL-15, which could be indicative of IL-15/IL-15Ralpha complexes detected by ELISA However, we did not find any correlation between IL-15 and IL-6 which suggested a mechanism of IL-6 induction by IL-15Ralpha independently of soluble IL-15. Also IL-15 and TNFalpha were none correlated, probably because induction of TNFalpha by IL-15 takes place in the microenvironment of target cells depending on cell contact and it is not reflected in acellular synovial fluid. In our experiments, we found a well-defined group of patients being negative to TNFalpha; however these patients express high or medium levels of IL-15, only 3 patients showed both TNF alpha and IL-15 high. We did not find any correlation between other pairs of cytokines in this work (IL-15Ralpha/TNFalpha and TNFalpha/IL-6). Therefore, these results reinforce our suggestion that IL-15 reverse signaling through membrane IL-15, could take place in RA synovium as other inducing pathway of IL-6 [
24]. Nevertheless, it is important to take into account complexities of the IL-15/IL-15R system, three different functional forms of IL-15 have been identified: 1) the soluble cytokine, 2) IL-15R-independent membrane-bound IL-15 and 3) membrane-anchored IL-15 through IL-15Rα [
13,
20,
23]. Also, IL-15Rα may be secreted as a functional soluble molecule (s-IL-15Rα) and could form a complex with IL-15 [
19,
21]. Thus, these results raise another question about the role of different presentation forms and complexities of cytokines/cytokines receptor system highlighting the need of personalized medicine in treatment using the cytokines as a target.
Conclusions
In our current work, we found higher levels of TNFalpha, IL-6 and IL-15Ralpha in RA than OA, but not of IL-15 suggesting that IL-15 could be implicated in the OA pathology. Also, our results confirm heterogeneity of the inflammatory process assessed in RA patients where high inter-individual diversity expression patterns of IL-15Ralpha, IL-15, IL-6 and TNFalpha were observed. Despite this, we have identified 4 main clusters of concentration of these cytokines in SF suggesting the importance of identifying disease subgroups to personalize treatment. Finally, we found correlations between IL-15Ralpha-IL-6 and IL-15Ralpha-IL-15, but we did not find any correlation between other pairs of studied cytokines, highlighting a role for IL-15Ralpha in the RA pathology.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AS, AM, AC, and YR performed the study, analyzed, interpreted the data and drafted the paper; YM and RB performed the statistical analysis; ME and GG participated in coordination of experimental work and revised the paper. All authors read and approved the final manuscript.