Circulating mRNAs and miRNAs as candidate markers for the diagnosis and prognosis of prostate cancer

Abstract

Circulating nucleic acids are found in free form in body fluids and may serve as minimally invasive tools for cancer diagnosis and prognosis. Only a few studies have investigated the potential application of circulating mRNAs and microRNAs (miRNAs) in prostate cancer (PCa). The Cancer Genome Atlas (TCGA) database was used for an in silico analysis to identify circulating mRNA and miRNA as potential markers of PCa. A total of 2,267 genes and 49 miRNAs were differentially expressed between normal and tumor samples. The prediction analyses of target genes and integrative analysis of mRNA and miRNA expression revealed eleven genes and eight miRNAs which were validated by RT-qPCR in plasma samples from 102 untreated PCa patients and 50 cancer-free individuals. Two genes, OR51E2 and SIM2, and two miRNAs, miR-200c and miR-200b, showed significant association with PCa. Expression levels of these transcripts distinguished PCa patients from controls (67% sensitivity and 75% specificity). PCa patients and controls with prostate-specific antigen (PSA) ≤ 4.0 ng/mL were discriminated based on OR51E2 and SIM2 expression levels. The miR-200c expression showed association with Gleason score and miR-200b, with bone metastasis, bilateral tumor, and PSA > 10.0 ng/mL. The combination of circulating mRNA and miRNA was useful for the diagnosis and prognosis of PCa.

Fig 1. Heatmap showing gene and miRNA differential expression in tumor tissue samples (1: purple) compared with surrounding normal tissues (2: orange) from TCGA data.

a Global gene expression analysis revealed 2,267 differentially expressed genes (P ≤ 0.001 and FDR ≤ 0.001); b miRNA expression analysis revealed 49 miRNAs differentially expressed (P ≤ 0.001 and FDR ≤ 0.001). Each line in the left axis represents a gene/miRNA. Columns represent each sample.

Fig 2. Flowchart of the analyses to define candidate genes.

*Gene selected from Pubmed database.

Fig 3. Differential expression of circulating mRNAs and miRNAs in plasma samples from patients with PCa and CTR and their respective ROC curves.

a Expression analysis of patients versus controls. b Differential expression of SIM2 and OR51E2 genes in plasma samples from patients versus controls, both with PSA level ≤ 4.0 ng/mL and their respective ROC curves. c ROC curve referring to the score developed from the following transcripts: miR-200c, miR-200b, OR51E2, and SIM2 genes; sensitivity of 67%, specificity of 75%. ROC = receiver operating characteristic, PCa = prostate cancer, CTR = controls subjects, PSA = prostate-specific antigen, AUC = area under the curve. *Statistically significant difference (Student’s t-test P < 0.05).

Fig 4. Comparison between circulating miR-200b expression levels and clinic-pathological features.

The superior line represents the miR200b expression according to the: a. presence or absence of metastasis; b. unilateral or bilateral tumors; c. Low or high PSA levels; and, d. Gleason score ≤ 6 or ≥ 7. e-h. The ROC (Receiver Operating Characteristic) curve analysis for each histopathological feature. PCa = prostate cancer; mPCa = metastatic prostate cancer; AUC = area under the curve. *Statistically significant (Student’s t-test P < 0.05).