Comparison of single‑nucleus and single‑cell transcriptomes in hepatocellular carcinoma tissue

Abstract

Single‑nucleus RNA sequencing (snRNA‑seq) is a method used to analyze gene expression in cells for which isolation is complex, such as those in hepatocellular carcinoma (HCC) tissues. It constitutes an alternative to single‑cell RNA sequencing (scRNA‑seq) by analyzing the nucleus rather than the whole cell; however, whether it can completely replace scRNA‑seq in HCC remains to be clarified. In the present study, scRNA‑seq was compared with snRNA‑seq in tumor tissue obtained from patients with HCC, using the 10X Genomics Chromium platform. Seurat was also used to process the data and compare the differences between the two sequencing methods in identifying different cell types. In the present study, the transcriptomes of 14,349 single nuclei and 9,504 single cells were obtained from the aforementioned HCC tissue. A total of 21 discrete cell clusters, including hepatocytes, endothelial cells, fibroblasts, B cells, T cells, natural killer cells and macrophages were identified. Notably, a high number of hepatocytes were detected using snRNA‑seq, while an increased number of immunocytes were identified in the tumor microenvironment using scRNA‑seq. Results of the present study provided a comprehensive image of human HCC at a single‑cell resolution. Moreover, results of the present study further demonstrated that snRNA‑seq may be adequate in replacing scRNA‑seq in certain cases, and snRNA‑seq performs at an improved level in hepatocyte sequencing. Combined use of the two sequencing methods may contribute to the study of intercellular interactions.

Keywords: hepatocellular carcinoma; single‑cell RNA sequencing; single‑nucleus RNA sequencing.

Figure 1.

Integrated scRNA-seq and snRNA-seq datasets in HCC to identify and characterize cell types. (A) Vlnplot of the ribosome proportion of two sequencing methods. (B) Uniform Manifold Approximation and Projection visualization of 14,349 single nuclei and 9,504 single cells from the same HCC sample. Clustering was divided into 7 main cell types. (C) Vlnplot of the mitochondrial proportion of two sequencing methods. (D) Clustering was divided into 21 clusters. (E) Dotplot of marker genes expressed in T cells (CD3D and CD3E), macrophages/monocytes (C1QA and C1QB), B cells (CD79A and IGHG1), NK cells (NCAM1 and FCGR3A), endothelial cells (PECAM1 and PLVAP), hepatocytes (ALB and APOA1) and fibroblasts (SOD3 and ACTA2). snRNA-seq, single-nucleus RNA sequencing; scRNA-seq, single-cell RNA sequencing; HCC, hepatocellular carcinoma.

Figure 2.

Comparison of cell types and cell numbers identified using scRNA-seq and snRNA-seq. (A) Barplot of numbers of T cells, macrophages/monocytes, B cells, NK cells, endothelial cells, hepatocytes and fibroblasts identified using scRNA-seq and snRNA-seq. (B) Barplot of proportions of T cells, macrophages/monocytes, B cells, NK cells, endothelial cells, hepatocytes and fibroblasts identified using scRNA-seq and snRNA-seq. (C) Barplot of cell compositions identified using scRNA-seq and snRNA-seq. (D) Barplot of numbers of 0-20 clusters identified using scRNA-seq and snRNA-seq. (E) Barplot of proportions of 0-20 clusters identified using scRNA-seq and snRNA-seq. (F) Barplot of nFeatures of T cells, macrophages/monocytes, B cells, NK cells, endothelial cells, hepatocytes and fibroblasts identified using scRNA-seq and snRNA-seq. (G) Barplot of nCount of T cells, macrophages/monocytes, B cells, NK cells, endothelial cells, hepatocytes and fibroblasts identified using scRNA-seq and snRNA-seq. snRNA-seq, single-nucleus RNA sequencing; scRNA-seq, single-cell RNA sequencing; NK, natural killer.

Figure 3.

Comparison of numbers of subclusters in each cell type between scRNA-seq and snRNA-seq. (A) Barplot of numbers of subclusters in hepatocytes identified using scRNA-seq and snRNA-seq. (B) Barplot of numbers of subclusters in T cells identified using scRNA-seq and snRNA-seq. (C) Barplot of numbers of subclusters in B cells identified using scRNA-seq and snRNA-seq. (D) Barplot of numbers of subclusters in macrophages/monocytes identified using scRNA-seq and snRNA-seq. (E) Barplot of numbers of subclusters in NK cells identified using scRNA-seq and snRNA-seq. (F) Barplot of numbers of subclusters in fibroblasts identified using scRNA-seq and snRNA-seq. (G) Barplot of numbers of subclusters in endothelial cells identified using scRNA-seq and snRNA-seq. snRNA-seq, single-nucleus RNA sequencing; scRNA-seq, single-cell RNA sequencing; NK, natural killer.

Figure 4.

Subgroup analysis of hepatocytes. (A) Heatmap of the top 20 differentially expressed genes in hepatocyte clusters (cluster 0, 2, 4, 8, 9, 12, 15, 16, 19). (B) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 0. (C) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 2. (D) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 4. (E) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 8. (F) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 9. (G) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 12. (H) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 15. (I) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 16. (J) GO enrichment (biological process) results of differentially expressed genes in hepatocyte cluster 19. GO, Gene Ontology.

Figure 5.

Pseudotime analysis of hepatocyte subclusters. (A) Plot of trajectory colored with pseudotime. (B) Plot of trajectory colored with subclusters (cluster 0, 2, 4, 8, 9, 12, 15, 16, 19). (C) Pseudotime heatmap of differentially expressed genes (with high expression levels).

Figure 6.

Subgroup analysis of T cells. (A) Heatmap of the top 20 differentially expressed genes in T cell clusters (1, 3, 6, 11). (B) Vlnplot of marker genes expressed in four T cell clusters (C01 naive CD4⁺ T cells, C03 naive CD4⁺ T cells, CO6 effector CD8⁺ T cells and C11 depleted CD8⁺ T cells). (C) GO enrichment (biological process) results of differentially expressed genes in T cell cluster 1. (D) GO enrichment (biological process) results of differentially expressed genes in T cell clusters 3. (E) GO enrichment (biological process) results of differentially expressed genes in T cell clusters 6. (F) GO enrichment (biological process) results of differentially expressed genes in T cell clusters 11. GO, Gene Ontology.

Figure 7.

Interaction between hepatocytes and T cells. (A) Cell-cell interactions following further screening using mean >1. (B) Ligand-receptor interactions between hepatocytes and T cells (all, combined snRNA-seq with scRNA-seq). (C) Ligand-receptor interactions between hepatocytes and T cells (snRNA-seq only). (D) Ligand-receptor interactions between hepatocytes and T cells (scRNA-seq only). snRNA-seq, single-nucleus RNA sequencing; scRNA-seq, single-cell RNA sequencing.