Bacterial DNA activates the innate immune system via interactions with Toll-like receptor 9 (TLR9). This receptor recognizes CpG-oligodeoxynucleotides (CpG-ODNs) mimicking the CpG dinucleotides in certain sequence contexts characterizing this DNA. Most studies have shown increased osteoclast differentiation by TLR ligands. We found that activation of TLRs (specifically TLR4 and TLR9) in early osteoclast precursors results in inhibition of receptor activator of NF-kappaB ligand (RANKL)-induced osteoclast differentiation. Our objective is to identify the mechanism leading to this inhibitory effect of a TLR ligand. Since both RANKL-RANK and CpG-ODN-TLR9 interactions result in NF-kappaB activation, p38 and ERK phosphorylation, and TNF-alpha synthesis (all implicated in osteoclastogenesis), we hypothesized that CpG-ODN (but not RANKL) in addition induces the synthesis of an anti-osteoclastogenic factor. Control osteoclast precursors, and cells treated with RANKL, CpG-ODN, or their combination were studied using DNA arrays (GEArray Q Series Mouse NF-kappaB Signaling Pathway Gene Array, MM-016, SuperArray). We found a marked increase in the mRNA levels of the osteoclastogenesis inhibitor interleukin-12 (IL-12) in osteoclast precursors treated with CpG-ODN and CpG-ODN + RANKL. Northern and Western analyses, together with ELISA, confirmed the DNA array studies. In correlation with these findings, IL-12 inhibited RANKL-induced osteoclast differentiation and specific anti-IL-12-antibodies inhibited the anti-osteoclastogenic effect of CpG-ODN. In conclusion, activation of TLR9 by its ligand, CpG-ODN, results in synthesis and release of IL-12 opposing RANKL-induced osteoclast differentiation.