Single-Cell Transcriptome Sequencing Reveals Intra-Tumor Heterogeneity and Rare Stemness-Related Cell Subpopulation in Hepatocellular Carcinoma

Abstract

Introduction: Single-cell genomics has emerged as a powerful strategy to delineate the complex molecular landscapes of cancers, particularly HCC. It is a state-of-the-art technology which can unprecedentedly address intra-tumoral heterogeneity, instead of using the traditional bulk-cell methods, which undesirably mask genuine cell-level biological variations by adding up signals from individual cells. Methods: We employed the perfect combination of Fluidigm C1 single-cell capturing system and next-generation sequencing to interrogate the intra-tumoral heterogeneity based on a patient-derived HCC tumor xenograft (PDTX) model. Upon the generated single-cell transcriptome sequencing data on 139 unselected HCC single cells, we applied unsupervised hierarchical clustering, principal component analysis, t-distributed stochastic neighbor embedding, heatmap visualization, and inhouse developed algorithm to undertake discovery exploration. We confirmed the discovery findings on the same PDTX sample and other relevant HCC cell lines, using fluorescence-assisted cell sorting, liver cancer stem cell (CSC) marker-sorted transcriptome sequencing, and sphere formation assays. Results: Single-cell transcriptomic landscape revealed an intra-tumoral heterogeneity pattern, which could be categorized into 2 major cell lineages by EPCAM expression. Moreover, we observed a negative correlation between EPCAM and CD13 expressions, suggested by the apparently mutually exclusive pattern for their gene expression enrichment. We also identified a rare CD24+/ CD44+-enriched cell subpopulation residing within the tumor bulk. One unique feature of singlecell genomics approach is its ability to dissect the intricate inter-relationships among various liver CSC markers, of which adopting traditional cell-sorting method is usually limited by the practical combinations of cell markers. We were able to classify HCC single cells into different liver CSC groups and define their enrichment of individual liver CSC markers. HCC single cells could be exemplified into ten liver CSC groups, according to expressional enrichment pattern of liver CSC markers. We found the majority of HCC cells had enriched expression of liver CSC markers, ranging from one to three markers. Last but not least, upon confirmation of the single-cell analysis findings, we demonstrated CD24+/CD44+-enriched cells possess specific oncogenic gene expression signature and may represent a novel stemness-related subclone. CD24+/CD44+-enriched cells were detected in a panel of HCC cell lines but varied substantially in their proportions. Conclusion: Our proof-of-concept investigation has provided essential evidence supporting the importance of single-cell genomics in dissecting the tumor biology of HCC and liver CSCs. We anticipate single-cell genomics will provide not only useful insight to HCC research but also pragmatic guidance on better precision medicine in the long run.