PsA patients were recruited from the Rheumatology Department, St. Vincent’s University Hospital. Ethics for this study was approved by the St. Vincent’s University Hospital Ethics and Medical Research Committee and was performed in accordance with the Declaration of Helsinki. All patients gave fully informed written consent. PsA patients were defined according to CASPAR criteria. Baseline demographics of the PsA cohort are shown in Table 1 . Arthroscopies were performed under local anaesthetic using a Wolf 2.7 mm needle, and synovial tissue biopsies were obtained from the site of inflammation under direct visualisation as previously described (6). Biopsies were utilised for isolation of primary PsA synovial fibroblasts (PsAFLS). Ethics approval number RS18-055.

PsA synovial biopsies were digested with 1 mg/ml collagenase type 1 (Worthington Biochemical, Freehold, NJ, USA) in RPMI-1640 (Gibco-BRL, Paisley, UK) for 4 h at 37°C in humidified air with 5% CO₂. Dissociated cells were grown to confluence in RPMI-1640, 10% FBS (Gibco-BRL), 10 ml of 1 mmol/l HEPES (Gibco-BRL), penicillin (100 units/ml; Biosciences), streptomycin (100 units/ml; Biosciences) and fungizone (0.25 μg/ml; Biosciences) before passaging. Cells were used between passages 2–8.

PsAFLS were seeded in 6- (1x10⁵ cells/well), 48- (2x10⁴ cells/well) or 96- (2x10⁴ cells/well) well plates and allowed to attach overnight in RPMI-1640, 10% FBS, 10 ml of 1 mmol/l HEPES, penicillin, streptomycin and fungizone. Media was removed the following day and cells were serum starved by adding RPMI-1640 containing 1% FBS, 10 ml of 1 mmol/l HEPES, penicillin, streptomycin and fungizone for a further 24 h. PsAFLS were then pretreated with JAKi (Peficitinib (PEF), Filgotinib (FILGO), Baricitinib (BARI) (ACHEMBLOCK, CA,USA) and Upadacitinib (UPA) (Ambeed Inc, IL, USA); 5 μM) (JAKi were reconstituted in DMSO to 100 mM) or DMSO (5 μM; Sigma Aldrich) (vehicle control) for 1 h before being stimulated with Oncostatin M (OSM) (10 ng/ml; R&D) for 24 h. Concentration of 5 µM for all JAKi was used based on a previous study showing the dose response of all JAKi in FLS (29, 30). Additional experiments were performed to examine if JAKi alter secondary downstream effects of cytokines that do not signal through the JAK-STAT. Therefore, PsAFLS were plated in a 96- (2x10⁴ cells/well) well as outlined above and stimulated with IL-1β (10 ng/ml; Bio-Techne LTD, UK) +/- JAKi, with a DMSO control. Additional experiments were also performed for Tofacitinib (5 µM) under OSM (10 ng/ml) or IL-1β (10 ng/ml) stimulation.

To determine the effect of OSM on pSTAT3 expression, PsAFLS (1x10⁵ cells/well) were seeded in 6-well plates. Once confluent, cells were serum starved as previously described and stimulated with OSM (10 ng/ml) overnight, unstimulated (basal) PsAFLS were used as a control. Media was removed from the PsAFLS and ice-cold RIPA (Radio-Immunoprecipitation Assay) buffer (Sigma) containing 10 μg/ml phosphatase inhibitor cocktail and 10 μg/ml protease inhibitor cocktail (Sigma) was used to extract protein from the PsAFLS. Measurement of protein concentration was performed using a BCA assay (Pierce Chemical Co, Rockford, IL, USA). Protein (3 μg) was resolved by SDS-PAGE (5% stacking, 10% resolving), resolved proteins were then transferred onto nitrocellulose membranes (Amersham Biosciences, Buckinghamshire, UK) prior to 1 h blocking in wash buffer containing 5% non-fat milk with gentle agitation at room temperature. Membranes were incubated with rabbit polyclonal anti-pSTAT3 (Cell-Signaling Technology, UK), diluted in 5% non-fat milk containing 0.1% Tween 20 at 4°C overnight with gentle agitation. β-actin (1:5000, Sigma) was used as a loading control. Following three 15 min washes, membranes were incubated with appropriate horseradish peroxidase-conjugated secondary antibodies (1:5000) for 3 h at room temperature. The signal was detected using SuperSignal^® West Pico Chemiluminescent Substrate (Amersham Biosciences). Band densities were imaged using the ChemiDoc MP Imaging System (Bio-Rad, USA).

Supernatants from treated PsAFLS and DMSO control (5 μM) (2x10⁴ cells/well) seeded in 96-well plates were harvested and levels of IL-8, IL-6, and MCP-1 measured by specific ELISA (MCP-1: eBiosciences, USA, IL-8, IL-6; DuoSet ELISA, R&D systems, UK) according to manufacturer’s protocol.

To determine the effects of JAKi on specific genes in response to OSM stimulation, PsAFLS (1x10⁵ cells/well) were seeded in 6-well plates and stimulated as previously described. Total RNA was isolated using an RNeasy Plus mini kit (Qiagen, Germany) according to the manufacturer’s specifications. The integrity of the RNA samples was assessed using a bioanalyzer (Agilent, CA, USA). Samples with a 260:280 nm ratio of 1.8 or above were used in subsequent experiments. Total RNA (100 ng) was reverse transcribed to cDNA using a high capacity cDNA reverse transcription kit (Applied Biosystems, Cheshire, UK) and stored at -20°C until further use.

Gene expression data were quantified by RT-PCR using the QuantStudio 5 Thermal Cycler (Applied Biosystem, Lewes, UK). Reaction mixtures contained 1 μl of cDNA, SYBR green PCR mastermix (Applied Biosystems) and target mRNA specific primer pairs as follows: IL-6 for 5’ CCCTGAGAAAGGAGACATTGTAAC 3’, IL-6 rev 5’CCTCTTTGCTGCTTTCACACATG 3’, IL-8 for 5’ TTGGCAGCCTTCCTGATTTC 3’, IL-8 rev 5’ TGGCAAAACTGCACCTTCAC 3’, MCP-1 for 5’ GCTCGCTCAGCCAGATGCAA 3’, MCP-1 rev 5’ TGGTGAAGTTATAACAGCAGGTGA 3’, MMP1 for 5’ GCTAACAAATACTGGAGGTATGATG 3’, MMP1 rev 5’ ATTTTGGGATAACCTGGATCCATAG 3’, ICAM for 5′ AACCAGAGCCAGGAGACACTG 3′, ICAM rev 5′ GCGCCGGAAAGCTGTAGATG 3′.

Samples lacking multiscribe reverse transcriptase formed the negative controls to ensure target-specific quantification. Data were analysed using the comparative threshold cycle (Ct) method with normalization to the expression of RPLPO (for 5′ GCGTCCTCGTGGAAGTGACATCG 3′, rev 5′ TCAGGGATTGCCACGCAGGG 3′) and HPRT1 (for 5′ ATGGACAGGACTGAACGTCTTG 3′, rev 5′ GGCTACAATGTGATGGCCTC 3′) as endogenous controls.

Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR), reflecting oxidative phosphorylation and glycolysis, respectively, were measured using the Seahorse-XFe96 analyser (Seahorse Biosciences, UK). PsAFLS were seeded at 12x10³/well in a 96-well cell culture XFe microplate (Seahorse Biosciences) and allowed to adhere overnight. Following this, cells were treated with JAKi/DMSO (5 μM) for 1 h and then stimulated with OSM (10 ng/ml) for 24 h. Additional experiments were also performed in the presence of IL-1β (10 ng/ml). Basal oxidative phosphorylation/glycolysis were calculated by the average of five baseline OCR/ECAR measurements, respectively, obtained before injection of specific metabolic inhibitors; oligomycin (ATP-synthase-inhibitor) (2 μg/ml; Seahorse Biosciences), trifluorocarbonylcyanide phenylhydrazone (FCCP) (mitochondrial uncoupler) (5 μM; Seahorse Biosciences) and antimycin A (complex-III inhibitor) (2 μM; Seahorse Biosciences) and rotenone (2 μM; Sigma Aldrich). Oligomycin was injected to evaluate both the maximal glycolytic rate and ATP synthesis, determined by subtracting the amount of respiration left after oligomycin injection from baseline OCR. FCCP was injected to evaluate the maximal respiratory capacity (average of three measurements following injection). Maximal respiratory capacity was determined by subtracting baseline OCR from FCCP-induced OCR and the respiratory reserve (baseline OCR subtracted from maximal respiratory capacity).

PsAFLS (2x10⁴ cells/well) were seeded in 48-well plates for 24 h and serum starved as previously described. A single scratch wound was induced through the middle of each well with a sterile pipette tip and cells were subsequently treated with JAKi/DMSO (5 μM) for 1 h followed by stimulation with OSM (10 ng/ml) for 24 h.

PSA FLS migration across the wound margins was assessed and photographed using a phase-contrast microscope. Semi-quantitative analysis of cell repopulation of the wound was assessed. Briefly, cells were fixed with 4% paraformaldehyde, stained with 0.1% crystal violet and the number of migrating cells across the time zero margin was assessed.

BioCoat Matrigel™ Invasion Chambers (Becton Dickinson, UK) were used to assess PsAFLS invasion. Cells were seeded at 3x10⁴ cells/well in the migration chamber on 8 μM membranes pre-coated with matrigel. Cells were treated with JAKi/DMSO (5 μM) for 1 h and stimulated with OSM (10 ng/ml) for 48 h. Non-migrating cells were removed from the upper surface by gentle scrubbing. Migrating cells attached to the lower membrane were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. Cells from five random high-power fields for each well were counted to assess the average number of invading cells.

Supernatants from stimulated PsAFLS (2x10⁴ cells/well) seeded in 96-well plates were harvested for MMP1, MMP3 and MMP9 analysis by MSD assay (Meso Scale Diagnostics, USA) and MMP expression was measured according to manufacturer’s protocol.

Statistical analyses were performed using Prism 8 software. Wilcoxon Signed Rank test, one-way analysis of variance (ANOVA), Friedman Test with Dunn’s multiple comparison were utilised. p values of less than 0.05 (*p < 0.05), 0.01 (**p < 0.01), 0.001 (***p < 0.001) and 0.0001 (****p < 0.0001) were determined as statistically significant. All raw data are available on request.

As OSM was utilised to active the JAK-STAT pathway, initial experiments assessed the effect of OSM on pSTAT3 expression. Figure 1A demonstrates that OSM stimulates pSTAT3 in n=3 separate PsAFLS. To assess the impact of JAKi, we initially determined their effect on a range of pro-inflammatory mediators. Firstly, we stimulated the PsAFLS with OSM and found that MCP-1 and IL-6 secretion (both p < 0.05) were significantly increased following stimulation compared to control ( Figures 1B, C ). IL-6 gene expression was also significantly increased (p < 0.05), with an increasing trend observed for MCP-1 gene expression ( Figures 1B, C ). MCP-1 secretion was significantly reduced by Peficitinib, Upadacitinib (both p < 0.001) and Baricitinib (p < 0.05) ( Figure 1B ). Although not significant, Filgotinib also showed a strong decrease in MCP-1 secretion. In parallel, inhibition was also observed at gene level, with Baricitinib (p < 0.01) and Upadacitinib (p < 0.05) significantly decreasing MCP-1 mRNA expression ( Figure 1B ). Similarly, JAKi also reduced OSM-induced expression of IL-6 at both the protein and gene level ( Figure 1C ). Peficitinib (p < 0.001) and Upadacitinib (p < 0.01) displayed significant inhibition of IL-6 ( Figure 1C ). This observation was also observed at gene level with Baricitinib (p < 0.05) and Upadacitinib (p < 0.05) showing significant reductions in IL-6 expression. Although not significant, Peficitinib also displayed an inhibitory capacity on IL-6 gene expression ( Figure 1C ). In contrast to both MCP-1 and IL-6 expression, OSM significantly reduced IL-8 expression (p < 0.05) compared to control ( Figure 1D ). JAKi showed no significant effect on IL-8 secretion, however, there was an increasing trend observed for IL-8 mRNA expression ( Figure 1D ).

To examine whether the inhibitory effect of JAKi on pro-inflammatory mediators is paralleled by a shift in metabolism, we analysed the two major energy pathways: oxidative phosphorylation (OCR) and glycolysis (ECAR) in real time, using the Seahorse XFe- Analyser. Figure 2A displays the average bioenergetic profiles for ECAR and OCR of PsAFLS before and after injections of mitochondrial inhibitors: oligomycin, carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), antimycin A and rotenone in the presence of OSM. As shown in the seahorse profiles, stimulation with OSM increased the ECAR with no effect observed for OCR ( Figure 2A ). Quantification demonstrated a significant increase in baseline ECAR (p < 0.05) ( Figure 2B ) with no effect observed for baseline OCR ( Figure 2D ). Similarly, max glycolytic capacity (p < 0.05), but not max respiratory capacity was significantly increased by OSM stimulation ( Figures 2C, E ). Treatment with JAKi demonstrated a decrease in the ECAR bioenergetic profile for all JAKi, with minimal change observed for the OCR bioenergetic profile ( Figure 2A ). All JAKi reduced the OSM- induced glycolytic capacity of PsAFLS with a significant decrease in basal glycolytic capacity in response to Peficitinib (p < 0.01) and Baricitinib (p < 0.01) ( Figure 2B ), and a significant decrease in Maximal Glycolytic Capacity in response to Baricitinib (p < 0.0001) ( Figure 2C ). No change in basal respiration ( Figure 2D ), or maximal respiratory capacity in response to JAKi was observed ( Figure 2E ). This therefore resulted in a significant decrease in the ECAR : OCR ratio, signifying a shift away from glycolytic mechanisms and towards a reliance on mitochondrial respiration. This was most significant following treatment with Peficitinib (p < 0.01), Baricitinib (p < 0.05) and Upadacitinib (p < 0.05) ( Figure 2F ). Filgotinib also decreased both the glycolytic capacity and ECAR : OCR ratio, but this did not reach significance ( Figure 2F ). The impact of JAKi on PsAFLS bioenergetics is clearly demonstrated in the overall metabolic profile, whereby the glycolytic profiles of PsAFLS induced by OSM, shift towards a more quiescent state following treatment with JAKi ( Figure 2G ). To examine if JAKi alter secondary downstream effects of other cytokines that do not signal through the JAK-STAT pathway, we examined their effect on IL-1β-induced PsAFLS ( Figures 3 and 4 ). JAKi have no effect on IL-1β-induced MCP-1 secretion ( Figure 3A ). While some inhibitory effect was observed for IL-1β-induced IL-6 expression, this is not significant ( Figure 3B ). We also examined the bioenergetic profile of PsAFLS stimulated with IL-1β and the four JAKi ( Figures 3C–F ), in addition to Tofacitinib ( Figures 4A–C ). IL-1β increased the ECAR but not the OCR of the cells, with the JAKi’s including Tofacitinib showing minimal effect on either energetic pathways ( Figures 3C–F and Figures 4A–C ). Tofacitinib which predominantly inhibits JAK3 and to a lesser extent JAK2 significantly inhibited both MCP-1 and IL-6 in response to OSM ( Figure 4D ). Similar to the other JAKi, Tofacitinib had no effect on IL-1β-induced MCP-1 and IL-6 secretion ( Figure 4E ).

To further examine the effect of JAKi on PsAFLS pathogenic function, we examined the effect of JAKi on the invasive capacity of PsAFLS using Transwell Matrigel™ invasion chambers following stimulation with OSM. Representative images of PsAFLS invasion in unstimulated cells, OSM stimulated cells and OSM stimulated cells following treatment with JAKi are shown in Figure 5A . Quantitative analysis demonstrated the significant increase in invasive capacity following OSM stimulation (p < 0.05) compared to unstimulated cells ( Figure 5B ). This was significantly impacted by treatment with all JAKi, however Peficitinib (p < 0.001) and Filgotinib (p < 0.05) had the most significant reductions ( Figure 5B ). In parallel, we examined the effect of JAKi on the secretion of the cartilage destructive matrix metalloproteinase enzymes. Peficitinib significantly decreased MMP-1 (p < 0.05), MMP-3 (p < 0.05) and MMP-9 (p < 0.05). While a decrease for all three MMPs was also observed in Upadacitinib, Filgotinib and Baricitinib treated cells, this did not reach significance ( Figure 5C ). In response to the OSM induced MMP-1 expression, gene analysis showed similar decreasing trends, with Peficitinib (p < 0.01) and Baricitinib (p < 0.05) displaying the greatest inhibition. Although not significant, Filgotinib and Upadacitinib both reduced MMP-1 expression ( Figure 5D ).

Finally, the capacity of PsAFLS to migrate within the joint environment is associated with progressive and destructive joint disease, therefore, we next investigated the role of JAKi on migration of PsAFLS using a wound repair scratch assay. Figure 6A shows representative images demonstrating the increased migratory capacity of PsAFLS in response to OSM compared to unstimulated cells, in addition to the inhibitory effect of JAKi on PsAFLS migration, where repopulation of wound margins was inhibited by JAKi. Quantitative analysis demonstrated that OSM significantly induced migration of PsAFLS (p < 0.01) compared to basal control ( Figure 6B ). However, Peficitinib (p < 0.001), Filgotinib (p < 0.01) and Baricitinib (p < 0.05) all significantly decreased PsAFLS migration across the wound margins ( Figure 6B ). Although not significant, Upadacitinib also demonstrated strong inhibition of PsAFLS migration ( Figure 6B ). As migration of FLS is aided by adhesion molecules we also determined if JAKi influenced ICAM expression. Gene analysis showed a significant inhibition of OSM- induced ICAM expression by Upadacitinib (p < 0.01), all other inhibitors displayed decreases in ICAM expression, although these did not reach significance ( Figure 6C ).

A summary table outlining the main effects of each JAKi can be found in Table 2 .