The Identification of a Tumor Infiltration CD8+ T-Cell Gene Signature That Can Potentially Improve the Prognosis and Prediction of Immunization Responses in Papillary Renal Cell Carcinoma

Abstract

Background: CD8+ T cells, vital effectors pertaining to adaptive immunity, display close relationships to the immunization responses to kill tumor cells. Understanding the effect exerted by tumor infiltration CD8+ T cells in papillary renal cell carcinoma (papRCC) is critical for assessing the prognosis process and responses to immunization therapy in cases with this disease.

Materials and approaches: The single-cell transcriptome data of papRCC were used for screening CD8+ T-cell-correlated differentially expressed genes to achieve the following investigations. On that basis, a prognosis gene signature associated with tumor infiltration CD8+ T cell was built and verified with The Cancer Genome Atlas data set. Risk scores were determined for papRCC cases and categorized as high- or low-risk groups. The prognosis significance for risk scores was assessed with multiple-variate Cox investigation and Kaplan-Meier survival curves. In addition, the possible capability exhibited by the genetic profiles of cases to assess the response to immunization therapy was further explored.

Results: Six hundred twenty-one cell death-inhibiting RNA genes were screened using single-cell RNA sequencing. A gene signature consisting of seven genes (LYAR, YBX1, PNRC1, TCF25, MYL12B, MINOS1, and LINC01420) was then identified, and this collective was considered to be an independent prognosis indicator that could strongly assess overall survival in papRCC. In addition, the data allowed papRCC cases to fall to cohorts at high and low risks, exhibiting a wide range of clinically related features as well as different CD8+ T-cell immunization infiltration and immunization therapy responses.

Conclusions: Our work provides a possible explanation for the limited response of current immunization checkpoint-inhibiting elements for combating papRCC. Furthermore, the researchers built a novel genetic signature that was able to assess the prognosis and immunotherapeutic response of cases. This may also be considered as a promising therapeutic target for the disease.

Keywords: CD8+ T cell; papillary renal cell carcinoma; prognostic model; scRNA-seq; tumor infiltration immune cells.

Figure 1

Prognosis value of CD8+ T cells in papillary renal cell cancer (papRCC). (A) Heat map of multiple-variate Cox proportional risk model in terms of CD8+ T cell within papRCC. Z-score represents the risk score. (B) Kaplan–Meier survival curve investigation of CD8+ T cells in papRCC and ccRCC with TIMER (B), CIBERSORT (C), CIBERSORT-ABS (D), XCELL (E), and QUANTISEQ approaches.

Figure 2

Identification of papRCC tumor infiltration CD8+ T-cell-correlated genes. (A) View of a single-cell sample from a RCC case; annotated UMAP plots to identify a total of eight different cell types including epithelial cells, malignant cells, monocytes/macrophages, dendritic cells, CD8+ T, malignant cells, unknown cells, and endothelial cells. (B) Views of single cell from tumor-free, ccRCC paraneoplastic, ccRCC tumor, papRCC paraneoplastic, and papRCC tumor samples, respectively. (C) Violin plots to demonstrate CD8+ T cells. (D) Pie charts of the seven different cell types. (E) Bar graphs of the cell proportions of eight different cell types from tumor-free, ccRCC paraneoplastic, ccRCC tumor, papRCC paraneoplastic, and papRCC tumor samples, respectively. (F) Volcano plot of the differentially expressed genes (DEGs) in papRCC tumor infiltration CD8+ T cells. (G) Bar graph showed the results of KEGG pathway enrichment of DEGs in the papRCC tumor infiltration CD8+ T cells.

Figure 3

Construction of a CD8+ T-cell-correlated prognosis gene signature. (A) Volcano plot showing Cox regression investigation of survival-correlated papRCC-infiltrating CD8+ T-cell DEGs. (B) Forest plot lines of the top genes as screened by random survival forest investigation. (C) Violin plots showing the expression of the top genes in different cell types.

Figure 4

Expression levels of prognosis signature markers in the ccRCC- and papRCC-infiltrated CD8+ T cells. (A) Boxplots compare the gene expression of prognosis signature markers in the ccRCC- and papRCC-infiltrated CD8+ T cells. (B) Histograms compare the protein levels of prognosis signature markers in the ccRCC- and papRCC-infiltrated CD8+ T cells. *p<0.05 and **p<0.01.

Figure 5

Survival analysis based on the single signature gene. Kaplan–Meier curves show the risk definition of TCF25 (A), MINOS1 (B), MYL12B (C), RNF115 (D), LINC01420 (E), YBX1 (F), PNRC1 (G), and LYAR (H) in the papRCC (left column) and ccRCC (right column) TCGA samples.

Figure 6

Validation of prognosis gene labels for papRCC and subtypes. Kaplan–Meier (KM) investigation of the risk group defined with CD8+ T-cell-correlated gene tags in the TCGA training data set for (A) the general papRCC, (B) type 1 papRCC, and (C) type 2 papRCC. KM investigation of the risk model for CD8+ T-cell-correlated gene labels in the TCGA testing data set for (D) the general papRCC, (E) type 1 papRCC, and (F) type 2 papRCC. Three- and 5-year receiver operating characteristic curves from the TCGA training data set for (G) the general papRCC, (H) type 1 papRCC, and (I) type 2 papRCC. Three- and 5-year ROC curves from the TCGA testing data set for (J) the general papRCC, (K) type 1 papRCC, and (L) type 2 papRCC. (M) KM investigation of the risk model for CD8+ T-cell-correlated gene labels in the TCGA data set of ccRCC. (N) Three- and 5-year ROC curves from the ccRCC TCGA data set.

Figure 7

Relationship between risk scores and clinically related characteristics. (A) Distribution of risk scores as assessed by age, sex, and survival status in the papRCC. (B) Risk score distributions for clinically related stage, pathological stage, and pathological T stage in the papRCC. (C) Distribution of risk scores as assessed by age, sex, and survival status in type 1 papRCC. (D) Risk score distributions for clinically related stage, pathological stage, and pathological T stage in type 1 papRCC. (E) Distribution of risk scores as assessed by age, sex, and survival status in type 2 papRCC. (F) Risk score distributions for clinically related stage, pathological stage, and pathological T stage in type 2 papRCC. (G) Multiple-variate Cox regression forest plots of risk scores and clinically related characteristics in the GSE2748 data set.

Figure 8

Relationship between risk groups and immune checkpoints. Gene expression of (A) CTLA4, LAG3, and PDCD1 and (B) PDCD1LG2, TIGIT, and HAVCR2 in the high-/low-risk groups of papRCC. Gene expression of (C) CTLA4, LAG3, and PDCD1 and (D) PDCD1LG2, TIGIT, and HAVCR2 in the high-/low-risk groups of type 1 papRCC. Gene expression of (E) CTLA4, LAG3, and PDCD1 and (F) PDCD1LG2, TIGIT, and HAVCR2 in the high-/low-risk groups of type 2 papRCC.

Figure 9

Workflow of the current study.