Osteoclasts are multinucleated giant cells with bone resorbing activity. We previously reported that the expression of the transcription factor NFAT2 (NFATc1) induced by receptor activator of NF-B ligand (RANKL) is essential for the formation of multinucleated cells. We subsequently identified L-Ser in the differentiation medium as necessary for the expression of NFAT2. Here we searched for serine analogs that antagonize the function of L-Ser and suppress the formation of osteoclasts in bone marrow as well as RAW264 cells. An analog thus identified, H-Ser(tBu)-OMe⅐HCl, appeared to suppress the production of 3-ketodihydrosphingosine by serine palmitoyltransferase, and the expression and localization of RANK, a cognate receptor of RANKL, in membrane lipid rafts was down-regulated in the analog-treated cells. The addition of lactosylceramide, however, rescued the osteoclastic formation. When administered in vivo, the analog significantly increased bone density in mice and prevented high bone turnover induced by treatment with soluble RANKL. These results demonstrate a close connection between the metabolism of L-Ser and bone remodeling and also the potential of the analog as a novel therapeutic tool for bone destruction.Osteoclasts are multinucleated giant cells with bone resorbing activity and play a key role in bone remodeling in conjunction with osteoblasts (1, 2). It is well established that receptor activator of NF-B ligand (RANKL) plays a central role in osteoclastogenesis through its cognate receptor activator of NF-B (RANK) (3, 4). We previously found, using an in vitro model of osteoclastogenesis consisting of mouse cells and recombinant RANKL, that the expression of the transcription factor NFAT2 (NFATc1) induced by stimulation with RANKL is essential for the formation of mature osteoclasts (5). Interestingly, culture at high cell density in the in vitro differentiation system blocked progression to the multinucleated (MN) 3 cell stage. Subsequently, a high cell density was found to cause a change in the composition/state of the culture medium accompanying downregulation of NFAT2 expression, and we eventually identified L-Ser as a key component for the phenomenon (6). Although the differentiation medium contained seven nonessential amino acids (L-Ala, L-Asn, L-Asp, L-Glu, L-Gly, L-Pro, and L-Ser), no other amino acid showed such a characteristic property. In fact, no osteoclasts formed when only L-Ser was depleted from the differentiation medium with dialyzed serum. In this regard, D-Ser was completely inert, and moreover, when D-Ser was added together with L-Ser, there was a suppressive effect on NFAT2 expression/osteoclastic formation, implying certain analog(s) to be useful for suppressing osteoclastogenesis through the down-regulation of NFAT2 expression. Unfortunately, however, D-Ser had a toxic effect in RAW264 cells.Here we systematically searched for analogs with more effective suppressive activity and less toxicity. We consequently identified a novel serine analog, H-Ser(tBu)-OMe⅐HCl (or Ot.-Buty...