Spatial maps of hepatocellular carcinoma transcriptomes highlight an unexplored landscape of heterogeneity and a novel gene signature for survival

Cancer Cell International

Yanhui Zhang¹  Zhiqiang Qiu¹  Runfen Cheng¹  Yuhong Guo¹  Baocun Sun²  Danfang Zhang³  Nan Zhao³  Xueyi Dong³  Fan Li³  Xiulan Zhao³  Xinchao Ban³ 

[1] Department of Pathology, Cancer Hospital of Tianjin Medical University, 300060, Tianjin, China;Department of Pathology, Tianjin Medical University, No. 22 Qixiangtai Road, Heping District, 300070, Tianjin, China;Department of Pathology, Cancer Hospital of Tianjin Medical University, 300060, Tianjin, China;Department of Pathology, General Hospital of Tianjin Medical University, 300052, Tianjin, China;Department of Pathology, Tianjin Medical University, No. 22 Qixiangtai Road, Heping District, 300070, Tianjin, China;Department of Pathology, General Hospital of Tianjin Medical University, 300052, Tianjin, China;

关键词: Hepatocellular carcinoma (HCC);    Heterogeneity;    Spatial transcriptomics (ST);    Satellite nodules;    Gene signature;   

DOI  :  10.1186/s12935-021-02430-9

来源: Springer

【 摘 要 】

BackgroundHepatocellular carcinoma (HCC) often presents with satellite nodules, rendering current curative treatments ineffective in many patients. The heterogeneity of HCC is a major challenge in personalized medicine. The emergence of spatial transcriptomics (ST) provides a powerful strategy for delineating the complex molecular landscapes of tumours.MethodsIn this study, the heterogeneity of tissue-wide gene expression in tumour and adjacent nonneoplastic tissues using ST technology were investigated. The transcriptomes of nearly 10,820 tissue regions and identified the main gene expression clusters and their specific marker genes (differentially expressed genes, DEGs) in patients were analysed. The DEGs were analysed from two perspectives. First, two distinct gene profiles were identified to be associated with satellite nodules and conducted a more comprehensive analysis of both gene profiles. Their clinical relevance in human HCC was validated with Kaplan–Meier (KM) Plotter. Second, DEGs were screened with The Cancer Genome Atlas (TCGA) database to divide the HCC cohort into high- and low-risk groups according to Cox analysis. HCC patients from the International Cancer Genome Consortium (ICGC) cohort were used for validation. KM analysis was used to compare the overall survival (OS) between the high- and low-risk groups. Univariate and multivariate Cox analyses were applied to determine the independent predictors for OS.ResultsNovel markers for the prediction of satellite nodules were identified and a tumour clusters-specific marker gene signature model (6 genes) for HCC prognosis was constructed.ConclusionThe establishment of marker gene profiles may be an important step towards an unbiased view of HCC, and the 6-gene signature can be used for prognostic prediction in HCC. This analysis will help us to clarify one of the possible sources of HCC heterogeneity and uncover pathogenic mechanisms and novel antitumour drug targets.

【 授权许可】

CC BY