Objective Cancer-associated cachexia is a devastating pathological disorder characterized by skeletal muscle wasting and fat storage depletion. Circular RNA, a newly discovered class of noncoding RNAs with important roles in regulating lipid metabolism, has not been fully understood in the pathology of cachexia. We aimed to identify circular RNAs that are upregulated in adipose tissues from cachectic patients and explore their function and mechanism in lipid metabolism. Methods Whole transcriptome RNA sequencing was used to screen for differentially expressed circRNAs. Quantitative reverse transcription PCR was applied to detect the expression level of circPTK2 in adipose tissues. The diagnostic value of circPTK2 was evaluated in adipose tissues from patients with and without cachexia. Then, function experiments in vitro and in vivo were performed to evaluate the effects of circPTK2 on lipolysis and adipogenesis. Mechanistically, luciferase reporter assay, RNA immunoprecipitation, and fluorescent in situ hybridization were performed to confirm the interaction between circPTK2 and miR-182-5p in adipocytes. Results We detected 66 differentially expressed circular RNA candidates and proved that circPTK2 was upregulated in adipose tissues from cachectic patients. Then we identified that circPTK2 was closely related to the pathological process of cachexia and could be used as a diagnostic marker. Mechanistically, circPTK2 bound competitively to miR-182-5p and abrogated the suppression on its target gene JAZF1, which finally led to promotion of lipolysis and inhibition of adipogenesis. In vivo experiments demonstrated that overexpression of circPTK2 inhibited adipogenesis and enhanced lipolysis. Conclusions Our findings reveal the novel role of circPTK2 in promoting lipolysis and reducing adipogenesis via a ceRNA mechanism and provide a potential diagnostic biomarker and therapeutic target for cancer-associated cachexia.

Backgrounds Cancer-associated cachexia (CAC) is a metabolic syndrome characterized by progressive depletion of adipose and muscle tissue that cannot be corrected by conventional nutritional therapy. Adipose tissue, an important form of energy storage, exhibits marked loss in the early stages of CAC, which affects quality of life and efficacy of chemotherapy. MicroRNAs (miRNAs) are a class of noncoding RNAs that widely exist in all kinds of eukaryotic cells and play regulatory roles in various biological processes. However, the role of miRNAs in adipose metabolism in CAC has rarely been reported. This study attempted to identify important miRNAs in adipose metabolism in CAC and explore their mechanism to identify a new predictive marker or therapeutic target for CAC-related adipose tissue loss (CAL). Methods In this study, miRNA sequencing was firstly used to identify differentially expressed miRNAs related to CAL and the reliability of the conclusions was verified in large population samples. Furthermore, functional experiments were performed by up and down regulating miR-410-3p in adipocytes. The binding of miR-410-3p to Insulin Receptor Substrate 1 (IRS-1) was verified by Luciferase reporter assay and functional experiments of IRS-1 were performed in adipocytes. Finally, the expression of miR-410-3p in serum exosomes was detected. Results miR-410-3p was selected as differentially expressed miRNA through screening and validation. Adipogenesis was suppressed in miR-410-3p upregulation experiment and increased in downregulation experiment. Luciferase reporter assay showed that miR-410-3p binds to 3′ non-coding region of IRS-1 and represses its expression and ultimately inhibits adipogenesis. miR-410-3p was highly expressed in serum exosomes of CAC patients, which was consistent with results in adipose tissue. Conclusions The expression of miR-410-3p was higher in subcutaneous adipose tissues and serum exosomes of CAC patients, which significantly inhibits adipogenesis and lipid accumulation. The study shows that miR-410-3p could downregulate IRS-1 and downstream adipose differentiation factors including C/EBP-a and PPAR-γ by targeting 3′ noncoding region.