Multiple highly methylated CpG sites as potential epigenetic markers for the diagnosis of prostate cancer

• Published in: Clinical epigenetics Vol. 17; no. 1; pp. 122 - 12
• Main Authors: Roperch, Jean-Pierre, Charbonnier, Guillaume, Figiel, Sandy, Lamb, Alastair, Mills, Ian, Hennion, Claude, Cancel-Tassin, Géraldine, Cussenot, Olivier
• Format: Journal Article
• Language: English
• Published: London BioMed Central 11.07.2025; BioMed Central Ltd; BMC
• Subjects: Aged; Anopheles; Antigens; Biomarkers, Tumor - genetics; Biomedical and Life Sciences; Biomedicine; Cancer; CpG Islands; Diagnosis; DNA Methylation; Enzymes; Epigenesis, Genetic; Epigenetic DNA markers; Epigenetic inheritance; Gene Expression Profiling - methods; Gene Expression Regulation, Neoplastic; Gene Function; Genes; Genetic markers; Genomics; Human Genetics; Humans; Male; Methylation; Methylation-specific PCR; Methylome sequencing; Middle Aged; Oncology, Experimental; Prostate cancer; Prostatectomy; Prostatic Neoplasms - diagnosis; Prostatic Neoplasms - genetics; RNA; Sequence Analysis, DNA - methods; Tumors; Methylome sequencing; Methylation-specific PCR; Prostate cancer; Epigenetic DNA markers
• ISSN: 1868-7083; 1868-7075; 1868-7083
• DOI: 10.1186/s13148-025-01930-z

Background Prostate cancer (PCa) remains the leading cause of cancer deaths in men. The prostate-specific antigen (PSA) test is widely used for PCa screening, but it lacks specificity and can lead to over-diagnosis and over-treatment. New, effective and affordable markers are therefore needed. Results Using enzymatic methyl sequencing (EM-Seq), methylation-specific PCR (MS-PCR), and transcriptomics including a spatial approach, we analyzed tumor and non-tumor samples from radical prostatectomy specimens. Comprehensive methylome was performed in 15 paired samples of prostate cancer and their adjacent non-tumor tissue by EM-Seq. From over 4-million differentially methylated CpG sites, we identified 66 CpGs sites representing eight genes: CLDN5 , GSTP1 , NBEAL2, PRICKLE2 , SALL3 , TAMALIN/GRASP , TJP2 , and TMEM106A which were hypermethylated in PCa tissues ( p -value < 0.0001), and were confirmed by MS-PCR. A very good correlation between EM-Seq and MS-PCR results was observed (Pearson’s correlation of 0.93). Differential expression of these candidate genes was analyzed first, using an Affymetrix RNA array dataset from a cohort of 68 non-tumor samples and 101 tumors with different aggressiveness patterns and, second, by in situ expression using Visium 10X spatial genomics transcriptomics on eight prostate tissue sections with different tumor grades and non-tumor glands. Lower expression level was found, using RNA arrays, in tumor compared to non-tumor tissues for six of the eight genes ( p  ≤ 0.0001) and in tumor glands with high aggressiveness compared to non-tumor glands ( p  <  0.0001) for the eight genes using in situ transcriptomics. Conclusions Our study identifies promising DNA methylation markers for the diagnosis of prostate cancer.