Despite the availability of several treatment options and surgical approaches to tackle osteoarthritis (OA), complications and side effects are still common among affected individuals [
1,
2,
21‐
23]. Among several alterations in OA, such as inflammatory and biomechanical properties of cartilage, the reduction in lubricative properties of the synovial fluid, as well as its constituents (i.e. hyaluronic acid (HA)) are important factors [
22,
24]. Preliminary observation and evaluation of conventional approaches, like intra-articular viscosupplemental injections of HA have shown some convincing benefit and anti-inflammatory potential over the past years in both animal and human subjects [
13,
20,
22,
24‐
27]. In this study, we examined the anti-inflammatory, pro-anabolic and anti-catabolic properties of high-MW HA on human derived OA chondrocytes in an established and optimized in vitro pro-inflammatory macrophage co-culture model. In vitro inflammatory models are useful preliminary approaches to evaluate pharmacological or other biological agents before being tested on animals or in other in vivo conditions. Commonly existing in vitro inflammation models includes lung [
28], intestine [
29], brain [
30], skin [
31], but for OA, no simple, reproducible and easy handle in vitro inflammation model does exist. To study the biological effects of HA on human OA chondrocytes, we first established and optimized an in vitro inflammatory macrophage-human OA chondrocyte co-culture model. We chose the pro-inflammatory M1 macrophage and human OA chondrocytes in vitro transwell co-culture to be most appropriate, as it closely resembles the permeable synovial joint. The successful isolation of primary macrophages is difficult and not very reliable. Hence, we adapted our protocol based on Daigenault et al., 2010 in obtaining differentiated and activated pro-inflammatory M1 macrophages from human monocyte-derived THP-1 cell line [
32]. We also characterized the successful differentiation process of the THP-1 cells to rM0 or pro-inflammatory M1 macrophages by flow cytometry and cytokine assay, utilizing the cell surface marker CD14 (Fig.
1a) and pro-inflammatory cytokines. We observed a much higher cell surface expression of CD14 in the resulting rM0 phenotype when THP-1 cells were treated with PMA, as it is described in literature [
33]. Further differentiation with IFN-γ and LPS lead to a decrease in this surface marker, indicating a favorable differentiation towards M1 (Fig.
1a). These are well described phenomena. Several studies have shown that activated M1 macrophages secrete pro-inflammatory cytokines, like IL-1ß, TNF-α, and IL-6 [
34,
35]. The presence of these cytokines is a commonly observed phenomenon in OA [
24,
36,
37] and does closely mimic OA conditions in our established inflammation model. To further confirm the differentiation status of M1 macrophages, we treated differentiated M1 macrophages with the anti-inflammatory glucocorticoid agent dexamethasone, which suppresses the inflammatory response [
38]. Indeed, we observed a significantly higher secretion of IL-1ß, TNF-α, and IL-6 from M1 macrophages that being further suppressed in the presence of dexamethasone, compared to resting M0 macrophages (Fig.
1c). It should also be noted that CD14 is a widely used marker for monocyte-macrophage differentiation [
33]. But still some studies report a variation in the response of PMA to THP-1, as there is an increased CD14 level [
39‐
41] or poor to no response [
42]., The production of pro-inflammatory cytokines was used to confirm the presence of differentiated and activated M1 macrophages. The above data presents a robust and repeatable inflammatory model, consisting of differentiated pro-inflammatory M1 macrophages. To closely simulate the permeable OA synovial joint, we have optimized co-culture conditions of differentiated M1 macrophages, and human derived OA chondrocytes, utilizing the Transwell™ system. Hyaluronic acid (HA) is a major component of extracellular matrix (ECM), which is present in almost all tissues [
20]. The different roles of HA are highly dependent on its molecular weight [
43]. In the healthy joint, HA is an essential synovial fluid constituent with a high-molecular-weight (>1000 kDa) and moderate ionicity [
44]. Besides anti-inflammatory characteristics (i.e. reduced prostaglandin release [
45] and reduced oxidative stress [
46]), it contributes to elasticity and viscosity of the synovial fluid, shock absorption and cartilage ECM structural and functional integrity [
22]. HA exists as a polymer of various chain lengths and different crosslinking, but previous experiments indicate that HA molecular weights of 700–6000 kDa are best suited for cartilage repair [
13,
44]. Despite the fact that low-MW (>10 kDa) HA has appropriate viscoelasticity, moisture retention, and adhesive properties, high-MW HA is used in clinical practice [
13]. Many different HA preparations are commercially available in Europe and the USA, derived from different sources, with different molecular weights, joint residence time and rheological properties [
47]. Due to the reasons explained above, we have evaluated the effect of high-MW HA on pro-inflammatory cytokine secretion and catabolic, as well as anabolic cartilage markers in cultured human OA chondrocytes, present in our established transwell co-culture (Fig.
2a). Our results have shown a significantly reduced secretion of pro-inflammatory IL-1ß and TNF-α cytokines, but not IL-6 in the HA-treated group, compared to the control group (Fig.
2b). Also, high-MW HA treatment of human OA chondrocytes in our co-culture system resulted in an increased gene expression of anabolic cartilage markers (i.e. COL2A1 and ACAN) and reduced catabolic cartilage markers (i.e. MMP13 and MMP3), compared to the control group (Fig.
3a). Chondrocytes tend to de-differentiate in prolonged 2D cultures, marked by fibroblastic morphology and the expression of collagen type I (COL1A1) [
48,
49]. In our transwell co-culture, we observed a significant reduction of COL1A1 gene expression (Fig.
3a) and a significant increase in COL2A1/COL1A1 differentiation index in high-MW HA-treated human OA chondrocytes, compared to the non-treated control group (Fig.
3b). Our study confirms the beneficial effects of high-MW HA in previously reported in vitro and in vivo studies. Further, this study confirmed and strengthened previously reported anti-inflammatory, anti–catabolic and pro-anabolic properties of high-MW HA in OA chondrocytes and cartilage [
13,
22,
24‐
27]. The role of HA and his fragments has been described for several biological processes (i.e. inflammation) and clinical conditions, but very limited information is available for HA in OA chondrocytes [
50]. There is increasing evidence that HA fragmentation through inflammatory induction of hyaluronidases leads to worsening conditions in OA [
44]. The inhibitory activity of high-MW HA on the secretion of pro-inflammatory cytokines (i.e. IL-1ß, TNF-α), MMP gene expression (i.e. MMP13 and MMP3) and no significant change in IL-6 expression, as observed in our study, is consistent with previously published results.