| Popis: |
Background: HBV-associated HCC has distinct molecular characteristics and worse outcomes compared to non-HBV HCC. We hypothesized that HBV infection might introduce targetable dependencies specific to HBV-associated HCC. Here, we demonstrate the utility of a computational strategy integrating genomic, network and survival analysis for identifying potential therapeutic targets in HBV-associated HCC. Methods: RNA-Seq and CRISPR dependency screening data for 22 HCC cell lines were obtained from the Cancer Dependency Map (DepMap). Cell lines were classified as HBV RNA+ if they harbored more than 40 HBV RNA reads, measured using GATK PathSeq software. Separately, whole genome CRISPRi dependency screening was performed in Hep3B and SNU-368 cells with doxycycline (dox)- inducible HBx expression. Genes where HBV RNA+ status (for DepMap) or HBx expression (for dox-inducible Hep3B/SNU-368) were significantly associated with negative dependency score were classified as HBV differential dependencies. To identify shared targets across datasets, these dependencies were evaluated using two network analysis techniques: the MCODE graph clustering algorithm and network propagation. Finally, RNA-Seq and survival data for 371 HCC cases were obtained from TCGA, and for each gene, a Cox multivariate model was used to evaluate whether the combination of decreased RNA-Seq expression and HBV RNA+ status was associated with increased survival. Results: 10 of 22 DepMap HCC cell lines were HBV RNA+. From DepMap and dox-inducible Hep3B/SNU-368 dependency screening data, 674 HBV differential dependencies were identified (Wilcoxon p < 0.05 for all). MCODE analysis identified enriched gene networks harboring multiple dependencies, including clusters associated with chromatin remodeling, the DNA damage response, and MAPK/ERK and Wnt signaling. Network propagation identified 236 enriched genes (empiric p < 0.005 for all), including the histone deacetylase HDAC1 and transcription factors ELK1, HNF4A and NRF1. Notably, HDAC1 was significant in both the MCODE and network propagation analyses, and in multivariate analysis of TCGA clinical data, decreased HDAC1 expression was associated with increased survival in HBV RNA+ HCC cases (HR for death 0.10, 95% CI 0.03-0.40, p = 0.0009). Conclusions: Measurement of HBV RNA in HCC cell lines identified a set of HBV- associated differential gene dependencies for which network analysis showed enrichment of several genes/gene clusters, including HDAC1, ELK1, HNF4A and NRF1. Survival analysis highlighted HDAC1 as a potential target where decreased expression was associated with increased survival in HBV RNA+ HCC cases in TCGA. Beyond nominating HDAC1, this study demonstrates the utility of a computational strategy integrating genomic, network and survival analysis for evaluating therapeutic targets in a virus-associated cancer. Citation Format: Huat Chye Lim, Rigney Turnham, Katherine Lo, Yeonjoo Hwang, John Gordan. Computational identification of targetable dependencies in hepatocellular carcinoma (HCC) associated with hepatitis B virus (HBV) replication [abstract]. In: Proceedings of the AACR Special Conference: Advances in the Pathogenesis and Molecular Therapies of Liver Cancer; 2022 May 5-8; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2022;28(17_Suppl):Abstract nr PR01. |
