Sex-Modulated TACSTD2 Methylation and Hormone-Receptor Axis Interactions in Glioblastoma: Insights Toward Mechanistic Understanding

Abstract

Sexual dimorphism in glioblastoma (GBM) outcomes is well-established but mechanistically poorly understood. The DNA methylation at promoter-proximal CpGs can modulate transcription-factor binding, and it may interact with sex-specific hormone-receptor signaling. Utilizing TCGA-GBM (n = 403) clinical, methylation, and expression data, we conducted sex-stratified survival analysis of tumor-associated calcium signal transducer 2 (TACSTD2) CpG methylation via cSurvival, complemented by MethSurv and SMART App analyses and independent validation in Chinese Glioma Genome Atlas (CGGA) (n = 123). We conducted regulatory motif analysis (FIMO), chromatin context mapping (UCSC/ENCODE), network analysis (STRING), transcription factor enrichment (EnrichR), and expression–methylation correlations. We validated protein expression using CPTAC proteomics, tissue localization using Human Protein Atlas immunohistochemistry, expression heterogeneity using GEPIA and TISCH2, and cross-cancer patterns. Findings were validated in the CGGA (n = 123). Systematic screening of 16 TACSTD2 CpG probes identified two sites with profound sex-methylation interactions: cg16080552 (chr1:58,577,527, 246 bp upstream of TSS, χ² = 38.1, p = 2.6 × 10⁻⁸) and cg27398499 (chr1:58,578,200, 674 bp upstream of TSS, χ² = 20.3, p = 1 × 10⁻⁴). Hypomethylated males showed worst outcomes (cg16080552: observed = 135 events, expected = 89.4, χ² contribution = 23.23), whereas hypermethylated females showed best survival. Methylation–expression correlation analysis revealed 14/16 probes significantly correlated with TACSTD2 expression (r = −0.25 to r = −0.57), with cg16080552 showing strong correlation (r = −0.5, p = 3.7 × 10⁻⁵) and cg27398499 weaker correlation (r = −0.19, p = 0.14). Methylation was independent of copy number variation (p = 0.57). FIMO analysis within ±200 bp of key CpGs identified 350 significant transcription factor binding sites (p < 1 × 10⁻⁴), including NR1H4::RXRA (p = 5.29 × 10⁻⁶), NR2F1 (p = 5.39 × 10⁻⁶), ESRRB (p = 9 × 10⁻⁶), ESRRA (p = 1.14 × 10⁻⁵), Spi1 (p = 2.72 × 10⁻⁷), and SPIB (p = 6.55 × 10⁻⁷). EnrichR multi-database validation confirmed hormone-receptor enrichment across ChEA 2022, ENCODE Consensus, JASPAR PWM, ARCHS4, and TRANSFAC. ChIP-Atlas analysis revealed >40 independent ChIP-seq experiments showing ESR1, AR, FOXA1, and GATA3 binding at the TACSTD2 locus in hormone-responsive epithelial models, establishing regulatory precedence. STRING network analysis positioned TACSTD2 within hormone receptor regulatory networks (AR, ESR1/2, PGR, RARA/RARB/RARG, RXRA/RXRB/RXRG, NCOA1-3, EP300/CREBBP). Gene ontology enrichment: hormone-mediated signaling (FDR = 1.0 × 10⁻¹⁷), steroid hormone signaling (FDR = 1.0 × 10⁻¹⁵), cellular response to hormone stimulus (FDR = 1.0 × 10⁻¹²), and retinoic acid receptor signaling (FDR = 1.0 × 10⁻¹⁰). GEPIA analysis revealed TACSTD2 expression heterogeneity in GBM (range 0–4.0 log₂(TPM + 1)), whereas normal brain showed consistently minimal expression. TISCH2 single-cell analysis confirmed sparse expression in specific cellular subpopulations with co-expression of ZNF transcription factors (r = 0.20–0.38). CPTAC proteomics demonstrated mRNA-protein correlation (r = 0.379, p = 6.8 × 10⁻⁵). Human Protein Atlas immunohistochemistry confirmed TROP2 protein absent in normal brain and glioma but strong membranous/cytoplasmic staining in epithelial cancers. CGGA validation confirmed prognostic significance (all grades p = 0.00097; Grade III p = 0.0075). These findings tentatively suggest that TACSTD2 methylation may contribute to sex-specific differences in GBM biology through hormone-receptor-mediated mechanisms and lineage plasticity. We hereby propose TACSTD2 methylation as a candidate for further investigation as a sex-specific biomarker, recognizing that GBM is a polygenic disease. Furthermore, TACSTD2 likely functions within a more extensive network of gene–gene and gene–therapy interactions. This analytical work underscores the potential importance of hormone- and sex-dependent genetic interactions in GBM for precision medicine applications. Mechanistic validation through functional studies remains absolutely necessary. The insights derived from computational and hypothesis-generating methodologies have the potential to inform future investigative endeavors focused on sex-specific therapeutic approaches. Additionally, they may contribute to the integration of TACSTD2 methylation status with other genomic biomarkers within multigenic prognostic frameworks.