Osteoclast formation is a complex process that is yet to be clearly defined. Osteoclasts differentiate from monocytic precursors to large multinuclear cells via the actions of two crucial cytokines: macrophage colony stimulating factor (M-CSF) and receptor activator of NFKB ligand (RANKL). These two cytokines bind to the osteoclast precursor cells, activating various down stream signalling pathways, inducing genes required for differentiation and for activation of osteoclasts. Exposure of monocytic precursors to M-CSF alone leads to differentiation into macrophages. Osteoclast differentiation was suppressed by granulocyte macrophage colony-stimulating factor (GM-CSF), resulting in mononuclear cells, lacking tartrate-resistant acid phosphatase (TRAP) and a bone resorptive phenotype. Further analysis determined GM-CSF dosage and temporal effects on osteoclast formation, where higher doses and earlier treatments of GM-CSF result in greater suppression of osteoclast formation. To understand the TRAP negative mononuclear cell phenotype, various osteoclast related markers and nuclear factors were tested using quantitative real-time PCR. GM-CSF suppressed the mRNA expression of osteoclast markers, including TRAP and cathepsin K (CTSK). CTSK is a cysteine protease, involved in osteoclast activity of bone resorption. Furthermore, GM-CSF down regulated the expression of critical osteoclast-related nuclear factors, including nuclear factor of activated T-cells, cytoplasmic (NFATcI), which has been identified as playing a critical role in osteoclast differentiation and ftinction in mice and to some extent in humans. The suppression of crucial osteoclast markers and transcription factors by GM-CSF indicated an overriding of the RANKL signal and possible switching of the cellular phenotype away from osteoclasts. To determine the cellular phenotype of GM-CSF driven cell differentiation, flow cytometry analysis was employed. As the cells visualised as dendritic cell like, CDIa, a dendritic cell surface marker, was selected for investigation. CDIa was highly expressed in GM-CSF, M-CSF and RANKL (GMR) treated cells and was absent in osteoclasts (M-CSF and RANKL treatment). The CDI a observations were indicative of GM-CSF overcoming the RANKL signal for osteoclastogenesis and directing differentiation to dendritic-like cells. To ftirther understand the osteoclastogenesis suppressive effect of GM-CSF, a 19,000 gene cDNA microarray assay was examined. The microarray experiment showed that the CC chemokine, monocyte chemotactic protein I (MCP-l), was profoundly repressed by GM-CSF. CC chemokines are chemoattractants that are induced during inflammation and recruit monocytes to the site of inflammation. MCP-l and other CC chemokines, RANTES (regulated on activation normal T cell expressed and secreted) and macrophage inflammatory protein I alpha (MIP I a) permitted formation of TRAP positive multinuclear cells in the absence of RANKL. However, these cells were negative for bone resorption. In the presence of RANKL, MCP-1 significantly increased the number of TRAP positive multinuclear bone resorbing osteoclasts (p= 5.7x 105, while RANTES and MIPI a mildly increased the number of bone resorbing TRAP positive multinuclear cells. Furthermore, CC chemokines, MCP-1, RANTES and MIP I a are all induced when authentic bone resorbing human osteoclasts differentiate from monocyte precursors in vitro following M-CSF-RANKL treatment. The addition of MCP- 1, RANTES or MIP I a appeared to reverse GM-CSF suppression of osteoclast formation, resulting in TRAP positive multinuclear cells. However, only MCP- I recovered the bone resorption phenotype, while other chemokines, RANTES or MTPIa did not. The cognate receptors for MCP-1, in particular, CCR2b and CCR4, were potently induced by RANKL (12.6 and 49-fold, p= 4.0x107 and 4.0x108, respectively), whereas the chemokine receptors for RANTES and MTP I a (CCR I and CCR5) were not regulated by RANKL. Chemokine treatment in the absence of RANKL also induced MCP- 1, RANTES and MIP I a. Unexpectedly, treatment with MCP-I in the absence of RANKL resulted in 458-fold induction of CCR4 (p I.0xI010), while RANTES treatment resulted in two fold repression (p= I .Ox ioj. Since CCR2b and CCR4 are cognate MCP-I receptors, these data support the existence of an MCP-I autocrine loop in human osteoclasts differentiated using RANKL. All three chemokines in the absence of RANKL can induce TRAP positive multinuclear cells that are negative for bone resorption. However, as MCP-I can significantly increase the number of osteoclast formation and recover the bone resorbing osteoclast phenotype from GM-CSF suppression, MCP-1 is the most potent chemokine involved in osteoclast formation. MCP-1 induced TRAP positive multinuclear cells were characterised and found to be positive for calcitonin receptor (CTR) and a number of other osteoclast markers, including NFATcI. As NFATcI is associated with osteoclast maturity in mice and has even been referred to as a master regulator of osteoclast differentiation and ftinction, a strong induction of NFATcI should theoretically allow bone resorption of MCP-l mediated TRAP positive multinuclear cells. Although great NFATcI mRNA induction and activated nuclear NFAT were observed, MCP-1 did not result in the formation of bone resorbing osteoclasts in the absence of RANKL. Despite the similar visual phenotype and expression of mature osteoclast markers TRAP and CTR when compared to osteoclasts, RANKL treatment was required for the MCP- I induced TRAP positive, CTR positive, multinuclear cells to possess bone resorption activity. This suggested that MCP-1 mediated TRAP positive multinuclear cells were primed for RANKL signal, to ftirther differentiate into authentic osteoclasts. The lack of bone resorption was ftirther correlated with a deficiency in expression of certain genes related to bone resorption, such as CTSK and matrix metalloproteinase 9 (MMP9) and integrin aV. Another observation with implications for absence of the bone resorptive activity in MCP- I cell was the absence or disruption of the F-actin ring structure, correlating with the lack of integrin aV mRNA expression. It was hypothesised that as MCP-1 mediated TRAP positive multinuclear cells possessed a high induction of CTR, the addition of calcitonin would block multinucleation. Indeed, the exogenous calcitonin blocked the MCP-I induced formation of TRAP positive, CTR positive, multinuclear cells as well as bone resorption activity in the osteoclast controls, indicating that calcitonin acts at two stages of osteoclast differentiation in the human PBMC model. These data suggest that RANKL-induced chemokines are involved in osteoclast differentiation at the stage of multinucleation of osteoclast precursors and provides a rationale for increased osteoclast activity in inflammatory conditions where chemokines are abundant. Furthermore, MCP-I induced TRAP positive, CTR positive multinuclear cells appear to represent an arrested stage in osteoclast differentiation, afler NFATcI induction and cellular ftision, but prior to the development of bone resorption activity and therefore, could be termed 'preosteoclasts'.