bioRxiv2023
Despite advancements in early cancer detection and prevention methods, colorectal cancer (CRC) remains a significant global health problem. It is the third most common cancer and the second leading cause of cancer-related deaths worldwide. Additionally, there has been a marked increase of incidence in young adults, and the reasons for this tendency are not fully understood. Therefore, the need for more effective diagnostic methods of assessing disease risk at early stage is crucial. One of the newly developed blood-based circulating biomarkers with promising potential is the short hypermethylated region located at the Septin9 intronic region. Several clinical studies have proven its performance and applicability. However, the molecular mechanism behind this consistent and recurrent feature present in most of the CRC and related precancerous stages and why it is specific and advantageous for CRC development are poorly understood. Here, we used comprehensive epigenetic and gene expression profile analyses from different sources of human clinical samples and cell line data to link specific hypermethylation events at the Septin9 intronic loci, which initiate alternative transcription of the Septin9 gene. Through our investigation of TCGA-COAD RNA-seq samples (n=287), we found that there was no significant difference in global Septin9 levels between normal and tumor samples. However, we did observe a significant alteration in the transcript variant ratio between v1 and v2, suggesting the use of an alternative promoter. Our findings were further supported by our analysis of ATAC-seq data, which revealed that the v2 promoter conferred higher chromatin accessibility, which correlated with the expression of the v2 isoform. However, this was not supported by promoter or enhancer activity as measured by H3K27ac signals. Hypermethylation at the v2 promoter was confirmed in tumor samples, providing a possible explanation for the switch in variants. Protein sequence analysis confirmed small differences between Septin9 variant A(v1) and B(v2). However, AlphaFold2 indicates a substantial difference at the N terminus, which could impact protein phosphorylation. We hypothesize, that variant A (v1) and variant B (v2) are required for normal cell functions but shifting the balance towards v1 is more favourable for the tumor. Although very little is known about Septin9 and its function in CRC biology, we are confident that our study will help to emphasize the importance of understanding regulatory mechanisms behind tumor-specific biomarkers and helps to improve the application.