Sabine Vautier, Corentin Levacher, Florent Marguet, Edwige Kasper, Jean-Christophe Sabourin, Stéphanie Baert-Desurmont, Philippe Ruminy, Claude Houdayer
Our results do not suggest competition between the circRNAs and mRNAs of the key oncogenic genes that we investigated, but they do reveal differences in circRNA/mRNA expression patterns within normal and tumor tissues.
BACKGROUND/OBJECTIVES: Circular RNAs (circRNAs) are emerging regulators of genetic information that share the spliceosome biogenesis pathway with messenger RNAs (mRNAs), influencing their expression. There is an increasing body of evidence supporting their role in colorectal cancer (CRC) tumorigenesis. This study explored circRNAs in two distinct CRC tumorigenesis pathways: microsatellite instability (MSI) and microsatellite stability (MSS). We investigated competition between mRNA and circRNAs from their host genes, which could potentially disrupt normal gene regulation, and examined specific patterns of alterations in MSS and MSI tumors.
METHODS: Circular (circ) and linear (lin) exon-exon junctions were quantified using exon-specific probes targeting 48 genes involved in CRC predisposition and tumorigenesis processes. RNA was extracted from colorectal FFPE samples (stage 1 to 4 adenocarcinomas and adenomas). MSS tumors (MSS-TTs) and adjacent normal tissue (MSS-NT) were selected from 21 patients with a severe personal or family history of cancer. MSI tumors (MSI-TTs) and adjacent normal tissue (MSI-NT) were selected from 16 patients. Muscle content was also investigated as a potential confounding factor.
RESULTS: Principal component analysis distinguished NT from TT samples based on the sums of circular and linear junctions. CircRNA abundance was higher in samples with an elevated muscle content (Kruskal-Wallis test p-value = 0.034). Linear regression, adjusted for muscle content, showed significantly reduced global circRNA levels in tumors compared to in healthy tissues (MSI and MSS combined, p-value = 0.00268). In the MSS group, significant differences were observed between MSS-NT and MSS-TT in terms of circ/lin ratios and linear and circular junction counts for specific genes. MSI analysis revealed distinct gene profiles, with significant differences only in linear junction counts.
CONCLUSIONS: Our results do not suggest competition between the circRNAs and mRNAs of the key oncogenic genes that we investigated, but they do reveal differences in circRNA/mRNA expression patterns within normal and tumor tissues.