Document Type : Original Article
Authors
1
Student Research Committee, Babol University of Medical Sciences, Babol, Iran.
2
Department of Medical Biochemistry, Faculty of Medicine, Babol University of Medical Sciences, Babol, Iran.
3
Surgery Department, Rasoul Akram Hospital Clinical Research Development Center (RCRDC), Iran University of Medical Sciences, Tehran, Iran.
4
Department of Medical Biotechnology, Faculty of Medicine, Babol University of Medical Sciences, Babol, Iran.
5
Department of Biochemistry, Faculty of Medicine, Iran University of Medical Sciences, Tehran, Iran.
6
Department of Biostatistics and Epidemiology, School of Public Health, Babol University of Medical Sciences, Babol, Iran.
7
Department of Medical Genetics and Molecular Biology, Faculty of Medicine, Babol University of Medical Sciences, Babol, Iran.
8
Cellular and Molecular Biology Research Center, Health Research Institute, Babol University of Medical Sciences, Babol, Iran.
Abstract
Breast cancer (BC) remains the leading cause of cancer-related mortality in women
worldwide, primarily due to its high invasiveness and therapeutic resistance. This study
explores the role of noncoding RNAs, circular RNAs (circRNAs), in BC progression
through a competing endogenous RNA (ceRNA) network. Three GEO circRNA
microarray datasets (GSE101123, GSE165884, GSE182471) were retrieved,
normalized, and batch-corrected using ComBat. Differentially expressed circRNAs
(DEcircRNAs) were identified via limma (|log₂ FC| > 1, FDR < 0.05). Differentially
expressed miRNAs (DEmiRNAs) and mRNAs (DEgenes) were derived from TCGA-
BRCA RNA-Seq (1,091 tumors, 113 normals) and miRNA-Seq (1,078 tumors, 104
normals) data using DESeq2 (|log₂ FC| > 1, FDR < 0.05). CircRNAs harboring miRNA
response elements (MREs) were selected via CSCD, and miRNA-mRNA interactions
predicted through TarBase, prioritizing upregulated DEgenes. A ceRNA network was
constructed in Cytoscape based on expression concordance. The hsa_circ_0000378/hsa-
miR-205-5p/RAD51 axis was validated in 48 paired BC and adjacent non-tumor tissues
by RT-qPCR. Results indicated hsa_circ_0000378 upregulation (2.74-fold, p<0.001),
hsa-miR-205-5p downregulation (0.64-fold, p=0.0022), and RAD51 upregulation (3.46-
fold, p<0.001) in tumors. Spearman correlations showed negative associations between
hsa_circ_0000378 and hsa-miR-205-5p (r = -0.474, p<0.001), hsa-miR-205-5p and
RAD51 (r = -0.383, p<0.001), and positive between hsa_circ_0000378 and RAD51 (r =
0.497, p<0.001), supporting ceRNA regulation. ROC analysis revealed RAD51's
diagnostic potential (AUC=0.83, 95% CI: 0.74–0.90, sensitivity=0.81, specificity=0.55),
followed by hsa_circ_0000378 (AUC=0.75, 95% CI: 0.65–0.85, sensitivity=0.71,
specificity=0.77), and hsa-miR-205-5p (AUC=0.66, 95% CI: 0.56–0.76,
sensitivity=0.69, specificity=0.55). These results propose the
hsa_circ_0000378/hsa_miR-205-5p/RAD51 axis as a potential biomarker; mechanistic
validation and larger cohorts are needed for clinical application.
Keywords