Background: Metastasis continues to be a leading cause of mortality in osteosarcoma (OS) among pediatric and young adult populations, with few effective therapeutic options available. Despite immunotherapy advancements, its efficacy in OS is hindered by an incomplete understanding of the immunosuppressive tumor microenvironment (TME). Methods: We utilized multiplex imaging mass cytometry and phenoplexing to characterize immune and stromal cell populations within the TME of a tissue microarray comprising 51 primary OS tumors. The prognostic significance of TME cell abundance and spatial cell–cell distance was evaluated using Kaplan–Meier and Cox regression analyses. To investigate macrophage functionality in vivo, we employed orthotopic xenograft mouse models by co-injecting THP-1-derived M0 or M2 macrophages with 143B OS cells to assess their impact on tumor growth and pulmonary metastasis. Mechanisms of macrophage-mediated metastasis were explored using Luminex, ELISA, and transwell migration assays. Results: Our results showed that macrophages dominated the TME, with M0 and M2 subtypes significantly outnumbering M1 macrophages (M1) and other myeloid cells. T cells and myeloid-derived suppressor cells (MDSC) were the second and third most abundant immune populations, respectively. Among stromal cells, endothelial cells predominated over fibroblasts. While individual immunosuppressive cell populations (M2, MDSC, and Treg) showed no direct correlation with clinical outcomes, the collective abundance of M2 and MDSC was significantly associated with reduced metastasis-free survival (MFS, p = 0.0244) and recurrence-free survival (RFS, p = 0.0040). Notably, closer spatial proximity between M2 macrophages and immunosuppressive cells (p = 0.0248) or Ki-67⁺ cells (p = 0.0321) correlated with decreased MFS, suggesting the formation of an M2-centric immunosuppressive and pro-tumor hub. In vivo, co-injection of M2 macrophages with 143B cells significantly enhanced pulmonary metastasis (p = 0.0140). Luminex analysis identified M2-derived MIP-1α (CCL3) as a candidate chemokine driving OS cell metastatic potential. Conclusions: This study provides a high-resolution map of the OS TME, highlighting the prognostic significance of M2 and immunosuppressive cell interactions in driving metastasis, potentially through MIP-1α signaling. These findings establish a foundation for developing targeted immunotherapies to improve outcomes in metastatic OS.

Background: Osteosarcoma (OS) is a fast-growing malignant bone tumor that occurs most often in children and teenagers. Development of pulmonary metastasis is the primary cause of treatment failure and mortality. Our previous studies demonstrated that cytoplasmic p27 interacts with PAK1, enhancing PAK1 phosphorylation and promoting OS pulmonary metastasis. However, the cellular functions of p27 and PAK1 are primarily regulated by phosphorylation, and the roles of specific phosphorylation residues in modulating OS metastatic potential remain unclear. Methods: To study tumor invasiveness and lung metastasis, we employed a CRISPR-based knock-in method to introduce specific mutations—p27-T157A, p27-T157D, PAK1-T423E, and PAK1-K299R—into the 143B OS cell line, followed by in vitro invasion and orthotopic xenograft mouse experiments. These residues were selected for their therapeutic potential, as T157 regulates p27 nuclear–cytoplasmic shuttling, while T423 and K299 modulate PAK1 kinase activity. Results: No significant differences in pulmonary metastasis were observed across p27 mutants compared to parental controls. However, the p27-T157D mutant exhibited increased cytoplasmic mislocalization, elevated PAK1-S144 phosphorylation, and enhanced in vitro invasiveness compared to the p27-T157A mutant and parental 143B cells. The PAK1-K299R mutant, designed to be kinase-dead, showed negligible S144 phosphorylation, consistent with loss of kinase activity. Unexpectedly, this mutant displayed increased T423 phosphorylation and in vitro invasiveness, and significantly enhanced pulmonary metastasis in vivo compared to the PAK1-T423E mutant and parental controls. Conclusions: These findings highlight the complexity of targeting specific p27 and PAK1 phosphorylation sites as an anti-metastatic strategy for OS. While p27-T157 phosphorylation influences cytoplasmic localization and invasiveness, it does not significantly alter metastatic outcomes. Conversely, PAK1-T423 phosphorylation is critical in driving OS metastatic potential, and the kinase-dead K299R mutant’s unexpected pro-metastatic effect suggests that kinase-independent mechanisms or compensatory pathways may contribute to metastasis. Our findings suggest the necessity for a more comprehensive understanding of the phosphorylation dynamics of p27 and PAK1 in metastatic OS. They also indicate that conventional kinase inhibition may be insufficient and underscore the potential benefits of alternative or combinatorial therapeutic strategies, such as targeting kinase-independent functions or other upstream kinases involved in these regulatory pathways.