Adalimumab

For 18 COVID-19 patients, one milliliter of whole blood taken at D0 was pretreated with different molecules for 6 h at 37°C, followed by stimulation with immune ligands on single lyophilized spheres (LyoSphere^TM, Qiagen), as described under Blood collection and cytokine assay. The molecules used were those commonly administered to COVID-19 patients (33 (link)–43 (link)): hydroxychloroquine (100 μM, Inresa), anti-IL6 Tocilizumab (100 μg/mL, RoActemra, Roche), methylprednisolone (20 μg/mL, Mylan), anti-TNFα Adalimumab (10 μg/mL, Humira, AbbVie), recombinant human IL-2 (6 ng/mL, Sigma), recombinant human IFN-alpha (100 ng/mL, Sigma) and Nivolumab (1 μg/mL, Opdivo, Bristol Myers Squibb).

Preosteoclasts, prepared like the BMDMs, were plated in 96-well plates (7,000 cells per well) in standard medium supplemented with 20 ng/ml M-CSF and 50 ng/ml Receptor Activator of Nuclear Factor Kappa-B Ligand (RANKL) (R&D Systems, Minneapolis, MN, USA). On the 3rd day (after 48 h), the medium was replaced by the CM of BMDM, supplemented with RANKL and M-CSF. Where indicated, we also added 2 μg/ml of neutralizing antibodies (Ab) against IL1β (Kineret, anakinra, SOBI, Stockholm, Sweden), IL6-receptor (Actemra, tocilizumab, Roche, San Francisco, CA, USA), or TNFα (Humira, adalimumab, AbbVie Inc., Chicago, IL, USA). These neutralizing Ab, effective in both humans and mice (15 (link)–19 (link)), are referred to as anti-IL1β, IL6, and TNFα Ab, respectively. On the 4th day, cells were stained using a TRAP kit (Sigma-Aldrich, St. Louis, MO, USA), and multinucleated (>3 nuclei) TRAP-positive cells were defined as osteoclasts. Images were acquired at an original magnification of × 4 (Evos FLC, Life Technologies, MS, USA). The number of osteoclasts and the total osteoclast area were measured using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Adalimumab (Abbvie, North Chicago, Illinois), golimumab (Janssen, Beerse, Belgium) and abatacept (Bristol Myers Squibb, New York) were obtained from the pharmacy of Bicêtre hospital, AP-HP (Le Kremlin Bicetre, France). Certolizumab pegol (CZP) and non-pegylated Certolizumal (CZNP) were provided by UCB, Brussels, Belgium. Peptides were purchased from Pepscan (Lelystad, The Netherlands).

SF cultures were established by explant growth of synovial tissues obtained by arthroscopic knee biopsies from patients without previous joint disease at elective arthroscopy for minor traumatic lesions, or patients with RA at the time of prosthetic replacement surgery. Patients signed a written informed consent, and the study was approved by the Ethics Committee of Hospital 12 de Octubre, Madrid, Spain (N° CEI:17/085). All methods involving humans were performed in accordance with the relevant guidelines and regulations. SF were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% heat inactivated fetal bovine serum (FBS) (Lonza, Verviers, Belgium) and used after 3rd passage. For all tests SF lines were stimulated with either TNFα, IL6/sIL6R, or both together in DMEM 0.5% FBS, the time and dose of the treatment will be indicated for each experiment. TNFα, IL6 and sIL6R (PreproTech, Rocky Hill, NJ, USA) were reconstituted according to manufacturer instructions. Where indicated, cells were treated with inhibitors Actinomycin D (10 μg/ml) (Sigma-Aldrich Quimica SA, Madrid, Spain), Adalimumab (10 μM) (AbbVie, North Chicago, IL, USA), Ruxolitinib (1 μM) (Selleckchem, Houston, TX, USA) and Cycloheximide (5 μM) (Sigma-Aldrich Quimica SA).