Enrichment and identification of differentially expressed genes in hepatocellular carcinoma stem-like cells

Abstract

Cancer stem cells are considered to be tumor-initiating cells. To explain the initiation or progression of hepatocellular carcinoma (HCC), we previously established a culture system that may enrich hepatic cancer stem-like cells (HCSCs). However, the regulatory mechanisms by which HCSCs acquire stem cell properties remain unclear. In the present study, three pairs of HCSCs and case-matched human HCC cells were analyzed by high-throughput screening, and novel biomarkers and pathways for the regulation of HCSCs were identified. The results led to the identification and stratification of 406 differentially expressed genes (DEGs), among which 73 GO terms were found to be significantly associated with DEGs in HCSCs, and only complement and coagulation cascade pathways were identified during the development of HCSCs. By combining the results of the Gene Ontology and Kyoto Encyclopedia of Genes and Genomes analyses, it was revealed that 7 genes were downregulated in the complement and coagulation cascade pathways, and 7 miRNAs were predicted to target several downregulated genes involved in these pathways. The results may contribute toward hepatic cancer stem cell studies and novel drug research for HCC treatment.

Keywords: Hepatocellular carcinoma; cancer stem-like cells; differentially expressed genes; high-throughput screening; signaling pathway.

Figure 1.

Classification of raw reads in six samples. Quality assessment was performed to demonstrate the composition of raw reads in Hep3B, Huh7, MHCC97-H, Hep3B-C, Huh7-C and MHCC97-H-C samples. For example, Only Adaptor (N, M%) means that the number of reads containing adaptors is N, and the proportion is M% of the total reads.

Figure 2.

DEGs and their clustering map in 3 pairs of cell samples. (A) The identified DEGs are shown in a Venn diagram. In total, 406 DEGs overlapped in the 3 hepatic cancer stem-like cell lines compared with case-matched human hepatocellular carcinoma cells, among which 90 DEGs were upregulated and 316 were downregulated. (B) The clustering map shows all 406 significant DEGs identified in sphere-forming cells and their parental cells. Each column represents an experimental condition (e.g. Hep3B-C vs. Hep3B), and each row represents a gene. Expression differences are shown in different colors. Red corresponds to upregulation and green indicates downregulation. DEGs, differentially expressed genes.

Figure 3.

Gene ontology annotations for the DEGs of the predicted hepatic cancer stem-like cell biomarkers. The identified 406 DEGs were retrieved and annotated with WEGO software in three Gene Ontology domains: Molecular function, biological process and cellular component. The 7 most highly enriched terms for each domain are indicated in blue. DEGs, differentially expressed genes.

Figure 4.

Functional analysis of complement-related DEGs (differentially expressed genes). (A) The term subordination relationship of immune system process in GO (gene ontology) analysis. The subordination relationships of the GO terms are reflected by their positions in the figure. Corrected P-values are indicated in the figure. (B) Scatter plots of global gene expression patterns and 7 DEGs. The scatter plots describe global gene expression patterns between hepatic cancer stem cells and human hepatocellular carcinoma samples. Red, grey and green represent gene expression showing upregulated, equivalent- and downregulated levels. The positions of C1S, C1R, CFI, C3, SERPINA5, SERPING1 and KNG1 are indicated with blue dots. (C) Two-way hierarchical clustering showing the 7 complement activation-related genes. C1S, complement component 1s; CR1, complement component 1r; CFI, complement factor I; SERPINA5, serpin family A member 5; SERPING1, serpin family G member 1; KNG1, kininogen 1.

Figure 5.

Differentially expressed miRNAs involved in complement and coagulation cascade networks. Among the miRNAs that were differentially expressed according to miRNA high-throughput sequencing, it was predicted that hsa-miR-197 interacted with CFI, and SERPINA5 was targeted by hsa-miR-183, hsa-miR-338 and hsa-miR-450b. CFI, complement factor I; SERPINA5, serpin family A member 5.