Article title: Dissecting the heterogeneity of the microenvironment in primary and recurrent nasopharyngeal carcinomas using single-cell RNA sequencing
Authors: Peng W-S, Zhou X, Yan W-B, Li Y-J, Du C-R, Wang X-S, Shen C-Y, Wang Q-F, Ying H-M, Lu X-G, Xu T-T, Hu C-S
Journal: Oncoimmunology
Bibliometrics: Volume 11, Number 1, pages e2026583-1–e2026583-15
DOI: http://dx.doi.org/10.1080/2162402X.2022.2026583
When this article was originally first published online, all figures were unintentionally omitted. The figures that should have been included in this article are reproduced below:
Figure 1.
ScRNA-seq profiling of the primary and recurrent NPC tumor environments. (A) Schematic representation of the experimental strategy. (B) The t-SNE plot, showing celltypes (left), patient origin (middle), and pNPC or rNPC origin (right). (C) Canonical cell markers were used to label clusters by cell identity as represented in the tSNE plot. (D) Average proportion of 8 main celltypes of CD45+ immune cells among patients.
Figure 2.
The role of the T cells in the immune functions of pNPC and rNPC. (A) tSNE projection of 30,652 T/NK cells, showing the composition of 15 clusters (upper), pNPC and rNPC origin (lower left), main cell types (lower middle). Average proportion of main celltypes between pNPC and rNPC origin (lower right). (B) Canonical cell markers used to identify T/NK cell subtypes. (C) Cumulative distribution function showing the distribution of naïve (upper), cytotoxic (middle), and exhausted (lower) state scores in each CD8+ T cells subtype. A rightward shift of the curve indicates increased state scores. (D) Development trajectory of CD8+ T cells inferred by diffusion map, colored by cell subtypes. The inlet plot showed each cell with a pseudotime score from red to yellow, indicating early and terminal states, respectively (left). Heatmap showing the dynamic changes in gene expression along the pseudotime (right). (E) The cell density distribution of the pseudotime-ordered CD8+ T cells from pNPC and rNPC origin(upper). Violin plot indicating the expression of inhibitory and cytotoxic genes in CD8+ T cells from pNPC and rNPC samples. The p values were calculated by Student’s t test(lower). (F) Violin plot indicating the cytotoxic (left upper) or exhausted (left lower) score CD8+ T cells from pNPC and rNPC samples. Scatterplot showing the correlation between the cytotoxic and exhausted scores from pNPC and rNPC samples (right). (G) GSVA analysis of up-regulated pathways in CD8+ T cells, rNPC versus pNPC (upper). Heatmap showing the activity of TFs in CD8+ T cells from pNPC and rNPC samples (right). (H) Average proportion of each CD4+ T cells subtype between pNPC and rNPC origin. (I) Violin plots showed the IL2R (left), inhibitory (middle), and co-stimulatory (right) scores for CD4+ T cells from pNPC and rNPC samples. (J) GSEA shows top enriched pathways in rNPC-derived CD4+ T cells
Figure 3.
The role of the NK cells and B cells in the immune functions of pNPC and rNPC. (A) tSNE projection of NK cells, showing the composition of 2 main celltype(left) and pNPC/rNPC origin (right). (B) Expression of marker genes(left) and dotplot of canonical cell markers(middle) used to identify NK cell subtypes. Average proportion of main celltypes between pNPC/rNPC origin (right). (C) Violin plot indicating the cytotoxic (left upper), chemotaxis (left middle) and lectin-like receptor binding (left lower) score of NK cells from pNPC /rNPC samples. Violin plot indicating the expression of DEGs in NK cells from pNPC and rNPC samples. (D) Heatmap showing the activity of TFs in NK cells from pNPC and rNPC samples. (E) tSNE projection of B cells, showing the composition of 4 main celltype (left) and pNPC/rNPC origin (middle). Expression of marker genes used to identify B cell subtypes (right). (F) Average proportion of B cells subtype between pNPC and rNPC origin. (G) Scatterplot showing the DEGs between pNPC and rNPC-derived naïve(left) and plasma(right) B cells. (H) GSEA shows top enriched pathways in rNPC and pNPC-derived memory B cells.
Figure 4.
The role of the myeloid cells in the immune functions of pNPC and rNPC. (A) tSNE projection of myeloid cells, showing the composition of 9 clusters (left) and pNPC/rNPC origin (middle). Average proportion of main celltypes between pNPC and rNPC origin (right). (B) Canonical cell markers used to identify myeloid cell subtypes. (C) Cumulative distribution function showing the distribution of M1(left upper) and M2 (left lower) scores in each macrophage subtype. A rightward shift of the curve indicates increased state scores. Violin plot indicating the M1(right upper) and M2 (right lower) score of macrophages from pNPC /rNPC samples. (D) Heatmap showing the activity of TFs in NK cells from pNPC and rNPC samples. Violin plot indicating the expression of DEGs in macrophages from pNPC and rNPC samples. (E) GSVA analysis of enriched pathways in rNPC versus pNPC-derived macrophages. (F) Differences in pathway activities by GSVA among different DC subtypes (left). Percentage of mature DC subtype among pNPC and rNPC samples. (G) Scatterplot showing the upgraded DEGs between pNPC and rNPC-derived DCs (left). Violin plot indicating the immune-suppressive (right upper) or antigen-presenting (right lower) score of DCs from pNPC and rNPC samples. (H) GSVA analysis of enriched pathways in rNPC versus pNPC-derived DCs (left). Heatmap showing the activity of TFs in DCs from pNPC and rNPC samples (right). (I) RNA velocity plot of myeloid cells.
Figure 5.
Malignant epithelial cells profiles between pNPC and rNPC. (A) tSNE projection of epithelial cells, showing the 15 clusters (left), patient (middle) and pNPC/rNPC group origin (right). (B) The heatmap of the relative expression density of genes on each chromosome between rNPC and pNPC epithelial cells. (C) Expression of certain cluster markers and corresponding pathway enrichment. (D) Canonical cell cluster markers on the tSNE (upper), corresponding GO and KEGG enrichment of pathways of certain clusters (lower). (E) GSEA shows top enriched pathways in rNPC and pNPC-derived malignant cells. (F) The top 3 significant PPI modules upgraded in rNPC-derived malignant cells identified by MCODE. Nodes indicate variables, with redder color indicating greater node degree (number of connections).
Figure 6.
Immune infiltrating associated with patient prognosis and cell communication. (A) Heatmap of the normalized immune abundance and clinical parameters in 113 NPC patients estimated by CIBERSORTx. Patients were clustered into four groups, representing different TME composition. (B) The correlation between estimated subpopulations and progression-free survival in 88 NPC patients, cox regression HR and 95% CI were shown (upper). Kaplan-Meier plot showing that patients with high abundance of Treg have shorter progression free survival, with high and low groups divided by the median value (lower). (C) Heatmap show number of potential ligand-receptor pairs between immune cell groups predicted by CellphoneDB. (D) Dot plot showed selected ligand-receptor interactions between Treg and immune cells, separated by primary (left) and recurrent (right) samples origin.
The corrected version of this article now includes all figures and citations inserted in appropriate places in the text.
The publishers would like to apologize for any inconvenience that these errors may have caused.