Gene-body DNA methylation of ONECUT2 predicts its expression and prostate cancer aggressiveness in needle biopsies

Abstract

Background: ONECUT2 is a lineage plasticity driver and therapeutic target in aggressive prostate cancer (PCa). This study investigated whether ONECUT2 gene-body DNA methylation regulates its expression and assessed its potential as a biomarker in clinical specimens.

Methods: We analyzed associations between ONECUT2 gene-body methylation, expression, and patient survival across multiple datasets. The effect of DNA methylation on ONECUT2 expression was tested in prostate cancer cell lines using a DNA methyltransferase inhibitor (DNMTi). Validation was further performed in needle biopsy samples by targeted bisulfite sequencing for DNA methylation and RT-PCR for gene expression.

Results: ONECUT2 expression strongly correlated with gene-body DNA methylation and patient survival in multiple datasets. DNMTi treatment confirmed this relationship in prostate cancer cells. In 208 biopsies from prostate cancer patients, hypermethylation of gene-body of ONECUT2 was linked to higher ONECUT2 expression and effectively distinguished tumor from adjacent normal tissue (p < 0.001 and AUC = 0.86). It also predicted aggressive features, including higher Gleason score (p = 0.01 and AUC = 0.68), advanced T stage (p = 0.04 and AUC = 0.65), seminal vesicle invasion (p = 0.0024 and AUC = 0.76), and lymph node involvement (p = 0.0005 and AUC = 0.80).

Conclusion: Assessing ONECUT2 gene-body methylation in biopsies may serve as a surrogate for ONECUT2 expression and provide predictive insights into disease progression before surgery. Furthermore, suppressing ONECUT2 through DNMTi treatment represents a potential therapeutic strategy for aggressive PCa.

Supplementary Information: The online version contains supplementary material available at 10.1186/s40364-026-00890-7.

Keywords: Aggressive prostate cancer; Biomarker; DNA methylation; Gene expression; Gene-body; Needle biopsy; ONECUT2; Prostate cancer; Therapeutic target.

Fig. 1

ONECUT2 expression in multiple public PCa datasets. (A) ONECUT2 expression in normal prostates and localized PCas from the TCGA PRAD dataset. (B) ONECUT2 expression in normal prostate, localized PCa and metastatic PCa from GSE3325, GSE6919, GSE21034, GSE35988 datasets. (C) ONECUT2 expression in treatment naïve PCa and castration resistant prostate cancer (CRPC) from GSE32269 and GSE70770 datasets. (D) ONECUT2 expression in adenocarcinoma prostate cancer (ADPC) and neuroendocrine prostate cancer (NEPC) from GSE32967, GSE66187, GSE77930, GSE104786, GSE126078 datasets and by Beltran et al. [29] (E) Kaplan-Meier plots of PSA recurrence-free survival of PCa patients stratified by ONECUT2 expression after prostatectomy using TCGA PRAD, GSE21034, GSE70770, and GSE94767 datasets

Fig. 2

The DNA methylation status of cg24771804 located in the ONECUT2 gene-body in PCa. (A) Map of the ONECUT2 gene including promoter and transcribed regions. The tick marks represent CpG sites covered by the Illumina EPICv1 array. (B) The beta value of cg24771804 from TCGA PRAD tumors and normal-adjacent prostate tissues. (C) The beta value of cg24771804 in normal prostate, localized PCa, and metastatic PCa from GSE73549, GSE157272, and GSE174613 datasets. (D) Kaplan-Meier plots of PSA recurrence-free survival of PCa patients with the beta value of cg24771804 after prostatectomy from TCGA PRAD, GSE83917, and GSE127985 datasets. (E) The correlation between ONECUT2 expression and cg24771804 DNA methylation in the ONECUT2 gene-body based on the TCGA PRAD, E-MTAB 6131, GSE83917, GSE107298, and GSE183015 datasets. Correlation coefficients of each cohort are plotted

Fig. 3

ONECUT2 expression shows a strong positive association with gene-body DNA methylation in prostate cancer cell lines. (A) Western blots of expression of ONECUT2 protein in RWPE1, PC3, C4-2B and LNCaP cells. β-actin was used as a loading control. (B) The beta value of two CpG sites (cg10835584, cg24771804) in the ONECUT2 gene-body in RWPE1, PC3, C4-2B and LNCaP cells. (C) The effect of DAC treatment on ONECUT2 protein expression level in RWPE1, PC3, and C4-2B cells. β-actin was used as a loading control. (D) Outline of the modified COBRA procedure. Figure was created using BioRender.com. (E) The effect of DAC treatment on the transcriptional expression of ONECUT2 in RWPE1 cells. (F) The effect of DAC treatment on the transcriptional expression of ONECUT2 and the DNA methylation levels of cg10835584 and cg24771804 in C4-2B and PC3 cells

Fig. 4

ONECUT2 gene-body DNA methylation based on targeted bisulfite sequencing and correlation with gene expression in needle biopsies. (A) The experimental outline of the targeted bisulfite sequencing approach. Figure was created using BioRender.com. (B) Heatmap of the ONECUT2 gene-body DNA methylation pattern based on DNA methylation status of 0–8 CpGs in the gene-body region. 7–8 CpGs P value labed as red represented most signifcant difference between normal and cancer. (C) Robust linear regression analysis comparing the relationship between ONECUT2 gene-body DNA mean methylation levels and mRNA relative expression compared to LNCaP in needle biopsies. These data were converted from raw data using a Yeo-Johnson transformation

Fig. 5

ONECUT2 gene-body DNA methylation and correlation with prostate cancer aggressiveness and AUC values in needle biopsies. (A) ONECUT2 gene-body DNA methylation based on the percentage of 7 or 8 methylated CpG sites and Receiver Operating Curve (ROC) analysis comparing normal prostate versus PCa. AUC values are indicated. (B) Gleason grade 1, 2 versus 3, 4, 5 (C) T2 versus T3 (D) Extra-prostatic extension (EPE) - negative versus positive (E) Seminal vesicle (SV) invasion - negative versus positive (F) Lymph node metastasis - negative versus positive