Osteoclasts are the exclusivebone-resorbing cells that have a central role in bone homeostasis as well as bone destruction in cancer and autoimmune disease. Both mouse and human genetic studies have clearly proven that receptor activator of NF-kappa B ligand (RANKL; encoded by the Tnfsf11 gene) and its receptor RANK are essential for osteoclastogenesis. Although there have been several reports on RANKL-independent osteoclastogenesis, previous studies have never provided in vivo evidence showing RANKL can be substituted by other molecules using RANKL-or RANK-deficient genetic backgrounds. Thus, to date, there is no clear evidence of RANKL-independent osteoclastogenesis and no molecule has ever been proven capable of inducing osteoclast differentiation more efficiently than RANKL. Recently, lysyl oxidase (LOX), the enzyme that mediates collagen cross-linking, has been shown to induce human osteoclasts in the absence of RANKL and has a stronger osteoclastogenic activity than RANKL. Here, we investigated the effect of LOX on osteoclast differentiation using RANKL-and RANK-deficient cells to strictly explore RANKL-independent osteoclastogenesis. CD14(+) human peripheral blood cells as well as osteoclast precursor cells derived from wild-type, RANKL-and RANK-deficient mice were treated with RANKL and/or LOX in short-term (3 days) or long-term (3 weeks) experimental settings. LOX treatment alone did not result in the formation of tartrate-resistant acid phosphatase (TRAP)(+) cells or resorption pits in either short-term or long-term culture. In combination with RANKL, long-term treatment with LOX synergistically promoted osteoclastogenesis in cells derived from wild-type mice; however, this was abrogated in RANKL-deficient cells. Long-term treatment with LOX stimulated RANKL expression in mouse bone marrow stromal cells via the production of reactive oxygen species (ROS). Furthermore, LOX injection failed to rescue the phenotype of RANKL-deficient mice. These results suggest that LOX has the ability to induce RANKL expression on stromal cells; however, it fails to substitute for RANKL in osteoclastogenesis. (C) 2016 American Society for Bone and Mineral Research.