Subtype-specific lncRNA ZFHX4-AS1 promotes invasion and metastasis of ovarian cancer by interacting with HuR RNA binding protein to stabilize its antisense gene ZFHX4

1. Background

Ovarian cancer (OvCa) is a heterogeneous disease, posing a significant challenge to women’s health globally. Currently, the biological function of long non-coding RNAs (lncRNAs) in different subtypes of OvCa remains poorly characterized. Among these subtypes, mesenchymal OvCa is the most aggressive subtype with unfavorable survival outcomes.

2. Major findings

We performed integrative network analysis to uncover the heterogeneity of lncRNAs in different OvCa subtypes. We identified ZFHX4-AS1 as a mesenchymal subtype-specific master regulator driving the epithelial-mesenchymal transition (EMT). Overexpression of ZFHX4-AS1 is a predictor of poor patient survival. Silencing ZFHX4-AS1 inhibits OvCa cell viability, migration, EMT, and stemness properties in vitro and reduces metastasis of OvCa in vivo. Mechanistically, by binding to the RNA binding protein (RBP) HuR, ZFHX4-AS1 promotes the interaction of HuR and mRNA of the target antisense gene ZFHX4, leading to enhanced ZFHX4 expression.

3. Codes for data analysis

• Drug_sensitivity_analysis.Rmd: Code for drug sensitivity analysis. Gene expression profiles and drug sensitivity data of OvCa cell lines were downloaded from https://discover.nci.nih.gov/rsconnect/cellminercdb/. The processed results can be found in CTRP_OV.RData file in demo_data subdirectory.

• Integrative_network_inference_and_master_regulator_analysis.Rmd: Code for developing the subtype-specific regulatory network using the TCGA-OV dataset. The processed results can be found in RTN_network.rda file in demo_data subdirectory.

• Pan_cancer_analysis_of_lncRNA.Rmd: Code for the pan-cancer analysis (survival and correlation analysis)of ZFHX4-AS1 and ZFHX4.

• TCGA_OV_data_preprocessing.Rmd: Code for the pre-processing of TCGA-OV dataset downloaded from the University of California Santa Cruz (UCSC) Xena data portal.

• Univariate_and_multivariate_Cox_analysis.Rmd: Code for uni-variate and multi-variate Cox analyzing and visualizing. The processed results can be found in Expr_OV.RData and RTN_network.rda files in demo_data subdirectory.