**2. Results**

### *2.1. YP Acts as an Anti-Inflammatory Agent in the Articular OA Cell System Model*

To evaluate the anti-inflammatory potential of YP in the mineralization and inflammatory processes involved in OA development and progression, a first screening was performed on THP-1 macrophages (THP-1 MOM). The inflammatory response of lipopolysaccharide (LPS)-induced THP-1 MOM was significantly reduced by pre-treatment with YP in a dose dependent manner (Figure S2a), as previously reported [25,26]. Additionally, the increased levels of TNF α production in calcium/phosphate (Ca/P) hydroxyapatite (HAP)-treated THP-1 MOM, confirmed the induction of a pro-inflammatory response [29], which was reduced by YP pre-treatment (Figure S2b). Cell proliferation assays were performed to confirm that tested HAP and YP did not a ffect THP-1 MOM cell viability (Figure S3). Based on these results, YP was further tested on a previously established articular OA cell system, consisting of human chondrocytes and synoviocytes primary cell cultures [30].

Human synoviocytes and chondrocytes primary cells were pre-treated with YP for 24 h followed by IL-1β (Figure 1) and HAP (Figure 2) stimulation. The e ffect of YP was determined by measuring gene expression of the inflammatory marker cyclo-oxygenase-2 (COX-2) and levels of IL-6 released into the cell culture media. Pre-treatment with YP followed by IL-1β stimulation resulted in a significant downregulation of COX-2 and decreased levels of IL-6 in both type of cells, relative to non-treated cells (Figure 1a,b). No cytotoxicity was observed in chondrocytes and synoviocytes, when treated with di fferent YP concentrations (Figure S4).

**Figure 1.** Amentadione (YP) reduces the levels of inflammatory markers in articular-derived cells stimulated with IL-1β (**<sup>a</sup>**,**b**). Primary chondrocytes and synoviocytes were pre-treated with 10 μM YP for 24 h, followed by stimulation with 10 ng/mL IL-1β during di fferent time points. (**a**) Relative gene

expression of the inflammatory marker COX-2 was determined by qPCR, at 3 h and 6h post IL-1β stimulation in chondrocytes and synoviocytes. (**b**) Levels of IL-6 in cell culture media 6 h post IL-1β stimulation, determined by ELISA. Cells treated with 2 μM dexamethasone (DXM) were used as a positive anti-inflammatory control. Data are presented as means of at least three independent experiments. All graphs show mean ± SD. One-way ANOVA and multiple comparisons were achieved with the Dunnett's test. Statistical significance was defined as *p* ≤ 0.05 (\*), *p* ≤ 0.005 (\*\*) and *p* ≤ 0.0005 (\*\*\*).

**Figure 2.** YP downregulates the inflammatory marker COX-2 in articular-derived cells stimulated with hydroxyapatite (HAP). Primary chondrocytes and synoviocytes were pre-treated with 10 μM YP for 24 h, followed by stimulation with 750 μg/mL HAP during 6 h. Relative gene expression of COX-2 was determined by qPCR, at 6 h post HAP stimulation in chondrocytes and synoviocytes. Data are presented as means of two independent experiments, with duplicates. All graphs show mean ± SD. One-way ANOVA and multiple comparisons were achieved with the Dunnett's test. Statistical significance was defined as *p* ≤ 0.05 (\*) and *p* ≤ 0.005 (\*\*).

Interestingly, a similar upregulation of the inflammatory marker COX-2 was observed in chondrocytes and synoviocytes treated with HAP, which was reduced by YP pre-treatment (Figure 2). No cytotoxicity was observed in chondrocytes and synoviocytes, when treated with different HAP concentrations (Figure S5).

These results demonstrate a promising anti-inflammatory effect of YP in the articular OA cell system model, through downregulation of inflammatory genes either when stimulated with IL-1β or treated with the mineralizing agen<sup>t</sup> HAP.

### *2.2. YP Modulates Cartilage Homeostasis under Mineralizing Conditions in an Ex-Vivo Cartilage Explant Model*

Since cartilage is the main affected tissue in OA, and ectopic mineralization is a known trigger of several joint alterations, including inflammation and cell differentiation, ultimately leading to cartilage degradation, cartilage tissue explants were first selected as an experimental model to evaluate ex vivo the effect of YP in response to HAP stimulation. Cartilage explants used in all experimental conditions were classified as normal- to early-OA tissues through the modified Mankin score [31]. Histological analysis revealed a smooth surface, a normal and uniform structural organization, and a normal to slight reduction in matrix staining, with Mankin total score ranging from 1 to 4 in the 1/13 modified Mankin scale (Table S1). The results showed that HAP treatment significantly upregulated collagen-10 (Col10), runt-related transcription factor-2 (Runx2) and matrix metalloproteinase-3 (MMP3) relative to control explants (Figure 3a), with simultaneous increased accumulation of MMP3 and the inflammatory marker IL-6 (Figure 3b). Pre-treatment of human cartilage explants with YP and further HAP stimulation, resulted not only in a significant down-regulation of the referred differentiation and ECM-related genes (Figure 3a), but also in decreased levels of the inflammatory marker IL-6 and the catabolic OA marker MMP3, responsible for ECM degradation (Figure 3b).

**Figure 3.** YP downregulates cell differentiation, extracellular matrix degradation and pro-inflammatory markers associated with osteoarthritis (OA) in the ex vivo cartilage explant model under HAP stimulation. Human cartilage explants were pre-treated with 10 μM YP for 24 h, followed by 72 h of 750 μg/mL HAP stimulation. Relative gene expression of Col10, Runx2 and MMP3 was determined by qPCR (**a**), and levels of MMP3 and IL-6 accumulation in the culture media were determined by ELISA (**b**). DXM indicates treatments with 2 μM dexamethasone. Data are presented as means of at least three independent experiments. All graphs show mean ±SD. One-way ANOVA and multiple comparisons were achieved with the Dunnett's test. Statistical significance was defined as *p* ≤ 0.05 (\*), *p* ≤ 0.005 (\*\*) and *p* ≤ 0.0005 (\*\*\*).

### *2.3. YP Function as a Protective Agent against Cartilage Deterioration under OA Promoting Conditions in an Explant-Based Co-Culture OA Model*

Since in the joint environment cartilage and synovial membrane are known to be involved in an interrelated and complex crosstalk affecting cartilage integrity and driving OA progression, an ex vivo explant-based co-culture OA model was developed and used to study the effects of YP in cartilage. Human cartilage explants were co-cultured with primary human synoviocytes and treated with IL-1β (Figure 4) and HAP (Figure 5) to simulate inflammatory and mineralizing conditions. Increased gene expression of COX-2, IL-6 and MMP3 in the co-culture cartilage explants treated with IL-1β, clearly indicated an induction of inflammatory reactions and ECM degradation at cartilage tissue level, which were consistently diminished in cartilage pre-treated with YP (Figure 4).

**Figure 4.** YP downregulates pro-inflammatory and ECM degradation markers associated with OA in the explant-based co-culture OA model under inflammatory stimulation with IL-1β. Cartilage explants co-cultured with human primary synoviocytes were pre-treated with 10 μM YP for 24 h, followed by 24 h of 10 ng/mL IL-1β stimulation. Relative gene expression of COX-2, IL-6 and MMP3 in cartilage explants were determined by qPCR. DXM indicates treatments with 2 μM dexamethasone. Data are presented as means of at least three independent experiments. One-way ANOVA and multiple comparisons were achieved with the Dunnett's test. All graphs show mean ±SD. Statistical significance was defined as *p* ≤ 0.005 (\*\*) and *p* ≤ 0.0005 (\*\*\*).

**Figure 5.** YP decreases the production of ECM degradation and pro-inflammatory markers in the explant-based co-culture OA model under mineralizing conditions. Cartilage explants co-cultured with human primary synoviocytes were pre-treated with 10 μM YP 24 h, followed by 72 h of 750 μg/mL HAP stimulation. Levels of MMP3 and IL-6 accumulation in the co-culture media were determined by ELISA. DXM indicates treatments with 2 μM dexamethasone. Data are presented as means of at least three independent experiments. One-way ANOVA and multiple comparisons were achieved with the Dunnett's test. All graphs show mean ±SD. Statistical significance was defined as *p* ≤ 0.05 (\*), *p* ≤ 0.005 (\*\*) and *p* ≤ 0.0005 (\*\*\*).

Additionally, increased levels of MMP3 and IL-6 in the cell culture media of co-culture cartilage explants treated with HAP demonstrated the interplay between mineralization and inflammation with consequent increased levels of inflammatory and ECM degrading markers, which were decreased with the YP pre-treatment (Figure 5).

Overall, considering the effects of YP at cartilage tissue level, evaluated using the cartilage explants and the explant-based co-culture models, the results sugges<sup>t</sup> that YP exerts a cartilage protective effect, by reducing inflammatory reactions and preventing chondrocyte differentiation towards extracellular matrix mineralization and degradation.

### *2.4. YP Downregulates NF-kB Expression and Inhibits Ikb*α *Phosphorylation in Primary Chondrocyte Cells*

Since YP was able to downregulate several pro-inflammatory mediators known to be directly regulated by the nuclear factor-κB (NF-kB) signaling pathway, we investigated whether the anti-inflammatory action of YP was due to its effect on NF-kB transcription and phosphorylation of its inhibitor IkB<sup>α</sup>.

In human primary articular chondrocytes, pre-treated with YP for 24 h followed by IL-1β stimulation, NF-kB expression was significantly downregulated at all-time points tested (Figure 6a). To determine the effect of YP in IkBα phosphorylation (pIkBα), known to precede NF-kB nuclear translocation, an initial experiment was performed to determine the optimal time point of pIkBα under IL-1β stimulation. Western blot analysis of chondrocyte protein extracts indicated increased levels of pIkBα from 30 min to 60 min of IL-1β treatment (Figure S6). Based on that, detection of pIkBα in chondrocytes pre-treated with YP for 24 h followed by 30 min IL-1β stimulation suggests a reduction of pIkBα in the YP treated chondrocytes relatively to the untreated and IL-1β stimulated cells (Figure 6b). Specific ELISA assays measuring pIkBα and total IkBα at 30 min shown that YP treatment reduces the ratio of pIkBα/total IkBα (Figure 6c), strongly indicating an effect of YP on IkBα phosphorylation.

**Figure 6.** YP downregulates NF-kB expression and inhibits IkBα phosphorylation in IL-1β-stimulated primary articular chondrocytes. (**a**) Relative gene expression of NF-kB was determined by qPCR at 3 h and 6 h post 10 ng/mL IL-1β stimulation. Data is presented as mean of three independent experiments. (**b**) Total protein extracts of chondrocytes cultured in untreated conditions, stimulated with 10 ng/mL IL-1β for 30 min, and pre-treated with YP (μM) followed by 30 min of 10 ng/mL IL-1β treatment,

were analyzed by Western blot to detect pIkB<sup>α</sup>. Position of relevant molecular mass marker (kDa) is indicated on the right side and GAPDH was used as loading control. (**c**) The pIkBα ratio (pIkBα/total IkBα) was determined in total protein extracts of chondrocytes cultured in control conditions (Ctr); 30 min of IL-1β (10 ng/mL) treatment; and pre-treated with YP for 24 h followed by 30 min of IL-1β stimulation (YP), by measuring the content of total and pIkBα with the specific InstantOne ELISA assay kit. Data is presented as mean of two out of four representative experiments. All graphs show mean ±SD. One-way ANOVA and multiple comparisons were achieved with the Dunnett's test. Statistical significance was defined as *p* ≤ 0.05 (\*), *p* ≤ 0.005 (\*\*) and *p* ≤ 0.0005 (\*\*\*).

These results demonstrate an anti-inflammatory effect of YP in articular chondrocytes, by downregulation of NF-kB expression, and inhibition of its activation through modulation of IkBα phosphorylation, and consequent downregulation of several NF-kB-related target genes.
