Back to Search View Original Cite This Article

Abstract

<title>Abstract</title> <p>Background Promoter hypermethylation recurrently silences tumor-suppressor and adhesion genes in squamous carcinomas, but the transcriptional regulators that oppose or permit this silencing in cervical cancer remain incompletely characterized. We combined network-based master regulator inference with orthogonal TCGA methylation-expression correlation to identify candidate upstream regulators associated with reduced methylation at silenced loci in cervical squamous cell carcinoma (CESC). Methods A literature-curated panel of 24 genes recurrently silenced by promoter methylation in cervical and head/neck squamous carcinomas was tested for transcription factor (TF) driver enrichment against a TCGA CESC ARACNe regulatory network (616 regulators tested, Benjamini-Hochberg corrected). Cross-context PubMed novelty scoring was applied to the top-ranked drivers in cervical and head/neck squamous cell carcinoma (HNSC) specifically. For the top candidate, Spearman correlation was computed between regulator RNA-seq expression and target promoter methylation (Illumina HM450) restricted to squamous cell carcinoma samples (n = 253) within the TCGA CESC cohort. To address the possibility that correlations reflect global tumor hypermethylation rather than locus-specific associations, partial Spearman correlations were computed controlling for a methylation burden index (mean HM450 beta across up to eight established CESC methylation markers: CDKN2A, RARB, DAPK1, CDH1, MGMT, MAL, SOCS1, PAX1). Results DACH1 showed the broadest panel coverage among ranked drivers (4 of 23 network-resolved genes: CADM1, EDNRB, ESR1, SLIT2; enrichment = 8.42, p-adj = 2.82e-3) and zero PubMed records specifically in cervical or head/neck squamous cell carcinoma, despite established roles as a methylation-silenced tumor suppressor in breast, gastric, and esophageal cancers. DACH1 expression correlated inversely with promoter methylation at three of four predicted targets in the squamous-restricted TCGA CESC cohort (n = 253): CADM1 (ρ = −0.249, p = 6.38e-5), ESR1 (ρ = −0.248, p = 6.60e-5), and SLIT2 (ρ = −0.263, p = 2.35e-5). EDNRB showed a nominally significant but below-threshold association (ρ = −0.157, p = 1.23e-2; |ρ| &lt; 0.2). DACH1 has previously been reported to repress ER-α (ESR1) signaling in breast cancer at the protein level; the promoter methylation correlation reported here for ESR1 in CESC is, to our knowledge, a distinct finding, whereas the CADM1 and SLIT2 associations appear to be entirely unreported in any cancer context. An identical exploratory test of SOX8, the next most comparable candidate, against its own predicted targets showed no supporting pattern, indicating the DACH1 signal is not a generic property of top-ranked drivers. Conclusions These convergent computational signals — network-derived regulon enrichment plus independent methylation correlation in an adequately powered cohort — nominate a DACH1-associated regulatory program at three loci in cervical squamous carcinoma that has not previously been reported in this tumor type. Partial correlation analysis controlling for global tumor methylation burden confirmed that these associations are locus-specific rather than attributable to a genome-wide hypermethylation phenotype. All findings remain correlative and require experimental validation. These results nominate DACH1 as a candidate epigenetic regulator of adhesion and signaling genes in cervical squamous carcinoma and provide a testable hypothesis for experimental follow-up, particularly ChIP-seq confirmation of DACH1 occupancy at the CADM1, ESR1, and SLIT2 promoters.</p>

Show More

Keywords

methylation squamous cervical carcinoma cesc

Related Articles

PORE

About

Connect