Integrative multi-omics reveals metabolic–stemness coupling and novel therapeutic targets in osteosarcoma chemoresistance

Chemoresistance remains the predominant cause of treatment failure in osteosarcoma (OS), yet its underlying molecular circuitry remains poorly defined. To address this gap, we conducted a series of mechanistic and clinically informed studies integrating molecular biology, epitranscriptomics, and single-cell transcriptomics to systematically decode resistance plasticity in OS. Our work demonstrates that dysregulation of ubiquitin-mediated proteostasis plays a central role in chemotherapy resistance. We identified the E3 ubiquitin ligase SOCS1 as a key suppressor of resistance, which is frequently downregulated in chemoresistant OS tissues and cisplatin-resistant cell models. Mechanistically, SOCS1 promotes K63-linked ubiquitination and proteasomal degradation of ACTN4. Loss of SOCS1 stabilizes ACTN4, thereby enhancing tumor stemness, self-renewal capacity, and tolerance to cisplatin. Restoration of SOCS1 expression or inhibition of ACTN4 reverses these phenotypes, establishing the SOCS1–ACTN4 axis as a druggable resistance driver with clinical relevance. In parallel, we uncovered an epitranscriptomic mechanism of metabolic adaptation in OS chemoresistance. Our findings demonstrate that METTL3-mediated m?A modification stabilizes the oncogenic long noncoding RNA LINC00520 in a YTHDF2-dependent manner. LINC00520 binds and stabilizes ENO1 by preventing FBXW7-mediated ubiquitination, leading to enhanced glycolytic flux and supporting cisplatin resistance under metabolic stress. Targeting this METTL3/LINC00520/ENO1 glycolytic axis suppresses tumor growth and restores drug sensitivity in vivo, highlighting an RNA modification–driven resistance pathway. To characterize resistance heterogeneity in the clinical setting, we performed single-cell RNA sequencing on chemotherapy-resistant OS patient samples collected at our center. We identified highly mutated malignant subpopulations of pre-osteoblast origin that emerged under chemotherapy selection. These resistant cell states demonstrated strong cell–cell communication with inflammatory cancer-associated fibroblasts (iCAFs) and exhausted CD8? T cells, particularly through MIF–CD74 and CXCL signaling, promoting immune evasion and survival under chemotherapeutic stress. These findings suggest that resistance maintenance is not solely tumor-intrinsic but also reinforced by tumor–stroma–immune crosstalk within the OS microenvironment. Collectively, these works establish a multi-dimensional model of chemoresistance in osteosarcoma driven by ubiquitination failure, epitranscriptomic reprogramming, metabolic adaptation, and immune remodeling. By decoding resistance as a coordinated and adaptive network rather than a single-pathway event, our findings provide a mechanistic foundation for multi-axis therapeutic strategies targeting protein stability, RNA modification, and microenvironmental signaling to overcome drug resistance in OS.