Hsa_circRNA_405498 and hsa_circRNA_100033 Serve as Potential Biomarkers for Differential Diagnosis of Type 1 Diabetes

Abstract

Abstract Context The role of circular RNAs (circRNAs) in type 1 diabetes (T1D) is largely unknown. Objective We aimed to identify some circRNAs as differential diagnostic biomarkers for T1D to distinguish between patients with latent autoimmune diabetes in adults (LADA) and type 2 diabetes (T2D). Methods The circRNA expression profiles were determined by Arraystar human circRNA microarray in T1D compared to controls (n = 6 each). The differentially expressed circRNAs were validated by real-time quantitative polymerase chain reaction using a validation cohort with 20 T1D and 20 controls. The diagnostic performances of the candidate circRNAs and the clinical parameters were assessed using the logistic least absolute shrinkage and selection operator (LASSO) regression model in a larger cohort with 457 individuals, including patients with T1D, T2D, and LADA, and controls. Results We identified 110 differentially expressed circular transcripts (53 upregulated and 57 downregulated) in T1D patients compared with controls. Further analysis showed that the levels of hsa_circRNA_405498 and hsa_circRNA_100033 were significantly downregulated in T1D compared to controls (both P < .05). Moreover, the expression levels of these 2 circRNAs showed sequential downregulation from controls, patients with T2D, LADA, to T1D (P < .05). The area under the curve (AUC) of receiver operating characteristic plots in logistic LASSO regression model showed high diagnostic accuracy for combination model with the 2 circRNAs and some clinical parameters in distinguishing T1D from LADA (AUC = 0.915), T2D (AUC = 0.993), and controls (AUC = 0.992). Conclusion Our study demonstrated that hsa_circRNA_405498 and hsa_circRNA_100033 are promising novel differential diagnostic biomarkers for T1D.