Background TNF-alpha (TNFa) has been shown to independently contribute to osteolysis in rheumatic diseases by promoting the generation of osteoclasts, the cells responsible for bone resorption. Both TNFa and RANKL, a key cytokine in normal osteoclast development, have been concomitantly detected in the synovial fluid of RA patients, and the neutralization of each by biologic therapeutics has proven to be efficacious in limiting bone loss. However, the role of TNFa in promoting human osteoclast differentiation and activity in the presence of RANKL is poorly understood.
Objectives In this study, we sought to determine the contribution of TNFa to RANKL-induced human osteoclast development and function under chronic conditions and to assess the effectiveness of various biologic agents in blocking the effect of TNF in this system.
Methods Primary human osteoclast precursors (OCP) were exposed to various combinations of M-CSF, RANKL and TNFa (100 ng/mL) for up to 7 days. Prior to addition to cells, the cytokine cocktail was pre-incubated for 30 min. with one of the following biologic agents at increasing concentrations: adalimumab (ADA), etanercept (ETN) or abatacept (ABAT). Osteoclast differentiation was determined by the presence of large multinucleated cells positive for tartrate-resistant acid phosphatase (TRAP) and by TRAP5b activity. Resorptive activity was assessed by measuring the release of either the degradation products of plate-bound Europium-labeled collagen (Eu-col) or the cross-linked C-telopeptide of type I collagen (CTX-I) from human bone chips.
Results In the absence of biologics, the addition of TNFa to OCP cultures with exogenous RANKL promoted earlier differentiation (increased TRAcP5b activity by day 4) and enhanced osteoclast activity (increased CTX-I levels by day 7) compared to RANKL alone. Based on early time course assessments using Eu-col, TNFa enhanced the kinetics of osteoclast maturation by 13 hrs and maintained 3-fold higher levels of osteoclast activity for up to 122 hrs. Among the biologics, ADA restored the rate of osteoclast differentiation and activity to levels comparable to RANKL alone at concentrations 10-fold lower than ETN (0.74 µg/mL and 20 µg/mL, respectively). Moreover, ETN was unable to maintain its inhibitory effect even at the highest dose tested, while ABAT was ineffective in preventing TNF-mediated augmentation of osteoclast development at all concentrations tested. Interestingly, the levels of several TNFa-induced pro-osteoclastogenic chemokines were reduced more significantly by ADA than ETN within 72 hrs of OCP exposure to TNF, suggesting that ADA is more effective at reducing TNF signaling than ETN at this early stage of osteoclast development.
Conclusions Our findings demonstrate that TNFa can significantly enhance the kinetics of RANKL-induced osteoclast differentiation and activity. Moreover, under chronic in vitro exposure of OCP to TNF, ADA is more effective than ETN (or ABAT) in mitigating the pro-osteoclastogenic effects of TNF. Overall, our results demonstrate the central role of TNF in RA joint destruction and the sustained potency of ADA in preventing osteoclast development due to chronic TNF exposure.
Acknowledgements Authors acknowledge Drs. Jochen Salfeld, Neelufar Mozaffarian, Martin Okun and Philip Sugerman for their support and helpful scientific input.
Disclosure of Interest B. Harvey Shareholder of: AbbVie Inc., Employee of: AbbVie Inc., Z. Kaymakcalan Shareholder of: AbbVie Inc., Employee of: AbbVie Inc.