Overexpression of Circular PRMT1 Transcripts in Colorectal Adenocarcinoma Predicts Recurrence and Poor Overall Survival

Abstract

Colorectal cancer (CRC) is one of the most prevalent and deadly neoplasms globally; this fact puts emphasis on the need for accurate molecular biomarkers for early detection and accurate prognosis. Circular RNAs (circRNAs) have recently emerged as very promising cancer biomarkers. In this study, we thoroughly examined whether the expression levels of circular transcripts of the protein arginine methyltransferase 1 (PRMT1) gene can predict the prognosis of patients diagnosed with colorectal adenocarcinoma, the most frequent type of CRC. Hence, a highly sensitive quantitative PCR (qPCR) assay was developed and applied to quantify circ-PRMT1 expression in cDNAs from 210 primary colorectal adenocarcinoma tissue specimens and 86 paired normal colorectal mucosae. Extensive biostatistical analysis was then performed to assess the potential prognostic power of circ-PRMT1. Significant overexpression of this molecule was observed in colorectal adenocarcinoma tissue samples in contrast to their non-cancerous counterparts. Moreover, higher circ-PRMT1 expression was correlated with poorer disease-free survival (DFS) and worse overall survival (OS) in colorectal adenocarcinoma patients. Interestingly, multivariate Cox regression analysis revealed that the prognostic value of the expression of this circRNA does not depend on other established prognostic factors included in the prognostic model. Furthermore, the stratification of patients based on TNM staging revealed that higher circ-PRMT1 levels were significantly related to shorter DFS and OS intervals, particularly in patients with colorectal adenocarcinoma of TNM stage II or III. In summary, this original research study provides evidence that circ-PRMT1 overexpression represents a promising molecular biomarker of poor prognosis in colorectal adenocarcinoma, not depending on other established prognostic factors such as TNM staging.

Keywords: circular RNA (circRNA); colon cancer; molecular tumor markers; non-coding RNA; prognosis; prognostic biomarkers.

Figure 1

Violin plots showing the distributions of circ-PRMT1 expression levels in colorectal adenocarcinomas and non-cancerous colorectal tissues. The y-axis is on a log₁₀ scale.

Figure 2

A comparison of the circ-PRMT1 expression levels in 86 cancerous colorectal tissues vs. their normal counterparts. The p value was estimated by the Wilcoxon signed-rank test.

Figure 3

A receiver operating characteristic (ROC) curve illustrating the ability of circ-PRMT1 expression to efficiently distinguish colorectal adenocarcinoma from normal colorectal tissue. The p value was estimated by the Mann–Whitney U test. Abbreviations: AUC, area under the ROC curve; CI, confidence interval.

Figure 4

Kaplan–Meier survival curves for the disease-free survival (DFS) and overall survival (OS) of colorectal adenocarcinoma patients. Patients with neoplasms overexpressing circ-PRMT1 had significantly poorer DFS (A) and OS (B) than patients with tumors expressing low circ-PRMT1 levels. The p values were estimated by the Mantel–Cox (log-rank) test.

Figure 5

Kaplan–Meier survival curves for the disease-free survival (DFS) and overall survival (OS) of colorectal adenocarcinoma patients, stratified based on their tumor location. Patients with colon tumors that overexpress circ-PRMT1 had inferior DFS (A) and OS (B) than those with colon tumors with lower circ-PRMT1 levels. Similarly, colorectal adenocarcinoma patients with rectal tumors overexpressing circ-PRMT1 had poorer DFS (C) and OS (D) than those with rectal tumors showing lower expression of circ-PRMT1. The p values were estimated by the Mantel–Cox (log-rank) test.

Figure 6

Stratified Kaplan–Meier survival curves for the disease-free survival (DFS) and overall survival (OS) of colorectal adenocarcinoma patients with TNM stage II or III. circ-PRMT1 overexpression may be used in addition to TNM staging, particularly for these two TNM stages, to distinguish those patients with higher DFS (A) and OS (B) probabilities. The p values were estimated by the Mantel–Cox (log-rank) test.