Tumor mitochondrial oxidative phosphorylation stimulated by the nuclear receptor RORγ represents an effective therapeutic opportunity in osteosarcoma

Osteosarcoma (OS) is the most common malignant bone tumor with a poor prognosis. The treatment strategy has remained virtually unchanged over the past 40 years. Here, we show that the nuclear receptor RORγ may serve as a potential therapeutic target in osteosarcoma. OS exhibits a hyperactivated OXPHOS program, which fuels the carbon source to promote tumor progression. We found that RORγ is overexpressed in OS tumors and is linked to hyperactivated OXPHOS. RORγ induces the expression of PGC-1β and physically interacts with it to activate the OXPHOS program by upregulating the expression of respiratory chain component genes. Knockdown or pharmacological inhibition of RORγ strongly inhibits OXPHOS activation, downregulates mitochondrial functions and increases ROS production, which results in OS cell apoptosis and ferroptosis. RORγ inverse agonists strongly suppressed OS tumor growth and progression in multiple cell-based xenograft models and in chemotherapy-resistant, patient-derived xenograft (PDX) models and sensitized OS tumors to chemotherapy without obvious toxicity in mice. Taken together, our results indicate that RORγ is a critical regulator of the OXPHOS program in OS and provide a potential therapeutic strategy for this deadly disease. Total RNA was extracted from 143B and MG63 cells, which were treated with vehicle or the inverse agonist GSK805 or XY018 or SR2211 for 48 h. RNA-seq libraries from 2 µg total RNA were prepared by using Illumina Tru-Seq RNA Sample, according to the manufacturer’s instructions. RNA sequencing (RNAseq) analysis was performed by BGI (Beijing Genomics Institute, Shenzhen, China). Briefly, mRNA with a polyA tail was enriched by magnetic beads with OligodT. cDNA was synthesized in a high-temperature system and purified by the kit, then using the BGISEQ platform for RNA-seq. To generate clean reads, raw reads were filtered by using SOAPnuke software. Then, we used HISAT (Hierarchical Indexing for Spliced Alignment of Transcripts) and Bowtie2 to compare clean reads and the reference genome sequence. RSEM was used to calculate the gene expression level of each sample. Transcripts were assembled, differentially expressed genes were generated, and heatmaps were drawn.