Single-Cell and Spatial Profiling of Tumor Microenvironment Heterogeneity in Human Osteosarcomas

Osteosarcoma, a highly heterogeneous malignant bone tumor, exhibits profound interpatient and intratumoral heterogeneity. Through integrated single-cell RNA sequencing and spatial transcriptomics, we identified distinct transcriptional profiles in the tumor microenvironment (TME) across patients, underscoring significant interpatient variation. Intratumorally, malignant cells predominantly diverged along two differentiation trajectories, converging into osteoblastic and chondroblastic functional states, potentially influenced by hypoxia. Notably, malignant cells exhibited stronger transcriptional clustering by sample-of-origin than by functional state, underscoring that intertumoral heterogeneity exceeds intratumoral heterogeneity. Spatial analysis revealed patient-specific cellular co-localization patterns within the TME, alongside conserved spatial relationships: vascular elements (endothelial cells, pericytes) demonstrated strong co-localization, while immune cells (T cells, myeloid cells) aggregated within shared regions. Crucially, these functional states occupied distinct microniches: osteoblastic regions were enriched with osteoclasts, vascular components, and immune cells, whereas chondroblastic regions displayed inverse cellular composition, preferentially localizing to hypoxic, vascular-poor niches and exhibiting significant enrichment of hypoxia-related signaling pathways. Furthermore, our data suggest tumor cells may activate fibroblasts via the SPP1 signaling pathway, positioning fibroblasts as key intermediaries in tumor-directed TME remodeling. This study systematically defines the landscape of osteosarcoma heterogeneity, delineating distinct functional cellular states, their spatially organized TME niches, and a putative SPP1-mediated tumor-fibroblast regulatory axis, thereby elucidating the functional and spatial architecture underpinning interpatient and intratumoral heterogeneity in osteosarcoma.