Tumor subtype-specific cancer-testis antigens as potential biomarkers and immunotherapeutic targets for cancers

Abstract

Cancer-testis (CT) antigens are potential targets for cancer immunotherapy because of their restricted expression in immune-privileged germ cells and various malignancies. Current application of CT-based immunotherapy has been focused on CT expression-rich tumors such as melanoma and lung cancers. In this study, we surveyed CT expression using The Cancer Genome Atlas (TCGA) datasets for ten common cancer types. We show that CT expression is specific and enriched within certain cancer molecular subtypes. For example, HORMAD1, CXorf61, ACTL8, and PRAME are highly enriched in the basal subtype of breast cancer; MAGE and CSAG are most frequently activated in the magnoid subtype of lung adenocarcinoma; and PRAME is highly upregulated in the ccB subtype of clear cell renal cell carcinoma. Analysis of CT gene expression and DNA methylation indicates that some CTs are regulated epigenetically, whereas others are controlled primarily by tissue- and subtype-specific transcription factors. Our results suggest that although for some CT expression is associated with patient outcome, not many are independent prognostic markers. Thus, CTs with shared expression pattern are heterogeneous molecules with distinct activation modes and functional properties in different cancers and cancer subtypes. These data suggest a cancer subtype-orientated application of CT antigen as biomarkers and immunotherapeutic targets.

Figure 1

CT gene expression frequency in human cancers. CT gene expression frequencies were calculated (see Methods) and genes having a >15% expression frequency in at least one tumor were plotted. Codes for cancer types: BRCA, breast; COAD, colon; GBM, glioblastoma; HNSC, head and neck; KIRC, kidney/renal clear cell carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; OV, ovarian; SKCM, skin cutaneous melanoma; and UCEC, endometrial carcinoma.

Figure 2

Cancer subtype-specific CT gene expression. Gene expression heatmaps of top cancer subtype-specific CT genes (see Methods) were generated using TCGA RNAseq log2(RPKM) values for (A) breast cancer, (C) lung adenocarcinomas, (D) colorectal cancer, and (E) renal clear cell carcinomas. (B), Gene expression heatmap from a microarray dataset (NKI-295) generated by cluster and TreeView software using medium removal of normalized data to validate basal-specific CT genes found in TCGA (A). Similarly, right panel of (C) is validation result from GSE26939 for TCGA lung adenocarcinoma results.

Figure 3

CT gene expression frequency in cancer subtypes. Cancer molecular subtypes were determined using consensus clustering (see Methods) and genes having a >30% overexpression frequency in at least one cancer subtype were plotted. For each cancer subtype, its percentage within the corresponding cancer was labeled in parentheses.

Figure 4

Correlation of CT gene expression to DNA methylation. (A) Plot of the least median Pearson correlation coefficients between gene expression (using logged RNAseq RSEM values) and DNA methylation (using Met450 M values) for CT-X genes (red) and non-X CT genes (blue) (see Methods). (B) Scatter plots of CXorf61 and PRAME gene expression (logged RSEM values) versus DNA methylation (beta values) in breast cancers and lung squamous cell carcinomas. Samples were labeled in colors to mark different cancer subtypes as shown in the legend. Infinium Met450 probes under study were marked on the figure, which can be used to locate exact chromosomal coordinates of the DNA methylation sites.

Figure 5

Prognostic value of CT genes in clear cell renal cell carcinomas (ccRCC). (A) Percentages of CT genes and all genes found to be prognostic in human cancers. (B) Histogram of gene density along survival analysis Coxph p values (in −log10) for eight tumors all using 200 randomly picked samples. (C) Kaplan-Meier analyses of patient outcome stratified by expression of SPANXC, C21orf99, and SSX1 genes in ccRCC (see Methods). (D) Expression of top poor prognostic CT genes in ccRCC subtypes. Heatmap was generated using cluster and TreeView software after medium removal using RNAseq RSEM values.