Cancer Testis explorer
1 May 2024
Package
CTexploreR 1.1.0
1 Introduction
Cancer-Testis (CT) genes, also called Cancer-Germline (CG), are a group of genes whose expression is normally restricted to the germline but that are found aberrantly activated in many types of cancers. These genes produce cancer specific antigens represent ideal targets for anti-cancer vaccines. Besides their interest in immunotherapy, they can also be used as cancer biomarkers and as target of anti-tumor therapies with limited side effect.
Many CT genes use DNA methylation as a primary mechanism of transcriptionnal regulation. This is another interesting point about CT genes as they represent suitable models to study DNA demethylation in cancer.
Currently the reference database gathering CT genes is the CTdatabase that was published in 2009, based on a literature screening(Almeida et al., 2009). This database is however not up to date. Recently identified CT genes are not referenced (in particular CT genes identified by omics methods that didn’t exist at the time) while some genes referred as CT genes appeared to be in reality expressed in many somatic tissues. Furthermore, the database is not in an easily importable format, some genes are not encoded properly (by synonyms names rather than by their official HGNC symbol names, or by a concatenation of both) resulting in poor interoperability for downstream analyses. More recent studies proposed other lists of CT genes like Wang’s CTatlas (Wang et al., 2016, Jamin et al., 2021, Carter et al., 2023). These lists were established using different criteria to define CT genes and hence differ substantially from each other. Moreover, these lists are usually provided as supplemental data files and are not strictly speaking databases. Finally, none of these studies describe the involvement of DNA methylation in the regulation of individual CT genes.
We therefore created CTexploreR, a Bioconductor R package, aimed to redefine
rigorously the list of CT genes based on publicly available RNAseq databases
and to summarize their main characteristics. We included methylation analyses
to classify these genes according to whether they are regulated or not by DNA
methylation. The package also offers tools to visualize CT genes expression and
promoter DNA methylation in normal and tumoral tissues. CTexploreR hence
represents an up-to-date reference database for CT genes and can be used as a
starting point for further investigations related to these genes.
2 Installation
To install the package:
if (!require("BiocManager")) {
install.packages("BiocManager")
}
BiocManager::install("CTexploreR")To install the package from GitHub:
if (!require("BiocManager")) {
install.packages("BiocManager")
}
BiocManager::install("UCLouvain-CBIO/CTexploreR")3 CT genes
The central element of CTexploreR is the list of
298 CT genes (see table below) selected based on
their expression in normal and tumoral tissues (selection details in the next
section). The table also summarises their main characteristics.
library(CTexploreR)## Loading required package: CTdatahead(CT_genes, 10)CTdata is the companion Package for CTexploreR and provides the omics
data that was necessary to select and characterize cancer testis genes as well
as exploring them. The data are served through the ExperimentHub infrastructure.
Currently available data are summarised in the table below and details can be
found in CTdata vignette or manuals.
CTdata()4 CT gene selection
In order to generate the list of CT genes, we followed a specific selection procedure (see figure below).
4.1 Testis-specific expression
Testis-specific genes (expressed exclusively in testis) and testis-preferential genes (expressed in a few somatic tissues at a level lower than 10x testis expression) were first selected using the GTEx database (Aguet et al., 2020).
Note that some genes were undetectable in the GTEx database due to multimapping
issues (these were flagged as “lowly expressed” in GTEX_category column).
A careful inspection of these genes showed that many of them are well-known
Cancer-Testis genes, belonging to gene families (MAGEA, SSX, CT45A, GAGE, …)
from which members have identical or nearly identical sequences. This is likely
the reason why these genes are not detected in the GTEx database, as GTEx
processing pipeline specifies that overlapping intervals between genes are
excluded from all genes for counting.For these genes, as testis-specificity
could not be assessed using GTEx database, RNAseq data from a set of normal
tissues were reprocessed in order to allow multimapping.
Expression of several genes became detectable in some tissues. Genes showing a
testis-specific expression (expression at least 10x higher in testis than in
any somatic tissues when multimapping was allowed) were selected, and flagged
as testis-specific in multimapping_analysis column.
The expression of the selected testis-specific or testis-preferential genes was further analysed in somatic cell types using scRNAseq data of normal tissues from the Human Protein Atlas (Uhlén et al., 2015). The aim was to ensure that the selected genes are not expressed in any specific somatic cell type, as the GTEx selection was based on bulk RNAseq data.
4.2 Germline-specific expression
We also used the single cell RNA-Seq data from the adult human testis transcriptional cell atlas (Guo et al., 2018) to ensure to select germline-specific genes but not genes that would be specific of any somatic cell type of the testis.
For each gene, the testis cell type has been determined as the cell type showing the highest mean expression for that gene. This allowed us to remove the genes for which this cell type corresponds to a testis somatic cell type (macrophage, endothelial, myoid, Sertoli or Leydig cells).
4.3 Activation in cancer cell lines and TCGA tumors
To assess activation in cancers, RNAseq data from cancer cell lines from CCLE (Barretina et al., 2012) and from TCGA cancer samples (Cancer Genome Atlas Research Network et al., 2013) were used. This allowed to select among testis-specific and testis-preferential genes those that are activated in cancers.
In the CCLE_category and TCGA_category columns, genes are tagged as
“activated” when they are highly expressed in at least one cancer cell line/
sample (TPM >= 10). However genes that were found to be expressed in all -or
almost all-cancer cell lines/samples were removed, as this probably reflects a
constitutive expression rather than a true activation. We filtered out genes
that were not completely repressed in at least 20 % of cancer cell lines/samples
(TPM <= 0.1).
4.4 IGV visualisation
All selected CT genes were visualised on IGV (Thorvaldsdóttir et al., 2013)
using a RNA-seq alignment from testis, to ensure that expression in testis
really corresponded to the canonical transcript. For some genes for which the
canonical transcript did not correspond to the transcript that we could see in
the testis sample, we manually modified the external_transcript_name
accordingly, to ensure that the TSS and the promoter region are correctly
defined. This is particularly important for methylation analysis that must be
focused on true promoter regions.
4.5 Regulation by methylation
Genes flagged as TRUE in regulated_by_methylation column correspond to
Genes that are significantly induced by a demethylating agent (RNAseq analysis of cell lines reated with DAC (5-Aza-2′-Deoxycytidine)).
Genes that have a highly methylated promoter in normal somatic tissues but less methylated in germ cells (WGBS analysis of a set of normal tissues).
For some genes showing a strong activation in cells treated with 5-Aza-2′-Deoxycytidine, methylation analysis was not possible due to multimapping issues. In this case, genes were still considered as regulated by methylation unless their promoter appeared unmethylated in somatic tissues or methylated in germ cells.
5 Available functions
For details about functions, see their respective manual pages. For all
functions, an option values_only can be set to TRUE in order to get the values
instead of the visualisation.
All expression visualisation functions can be used on all GTEx genes, not only on Cancer-Testis genes, as the data they refer to contains all genes.
5.1 Expression in normal healthy tissues
5.1.1 GTEX_expression()
Allows to visualise gene expression in GTEx tissues. We can for example see the difference of expression between testis-specific and testis-preferential genes. Testis-specific genes have been determined with a stricter specificity to the testis : they are lowly expressed in all somatic tissues and at least 10 times more in the testis. Whereas testis-preferential accepts a little expression outside the testis : they are lowly expressed in at least 75% of somatic tissues, but still 10 times more in the testis.
- Applied to testis-specific genes : we can clearly see the expression strictly limited to the testis. We can also see genes that are lowly expressed in GTEx, and have thus been characterized using multimapping (see below).
testis_specific <- dplyr::filter(
CT_genes,
testis_specificity == "testis_specific")
GTEX_expression(testis_specific$external_gene_name, units = "log_TPM")
- Applied to testis-preferential genes : we can see that the expression is less stringent to testis, as expected, with low expression in some other tissues, always with a strong testis signal.
testis_preferential <- dplyr::filter(
CT_genes, testis_specificity == "testis_preferential")
GTEX_expression(testis_preferential$external_gene_name, units = "log_TPM")
5.1.2 normal_tissue_expression_multimapping()
Allows to visualise the expression values obtained by counting or not multi-mapped reads in normal tissues. We can apply that to genes that are lowly expressed in GTEx (as seen above) in order to visualise their expression. First heatmap shows expression without multimapping while multimapping was allowed in the second one, where expression can be observed. This allowed their determination as testis-specific genes.
testis_specific_in_multimapping_analysis <-
dplyr::filter(CT_genes, lowly_expressed_in_GTEX)
normal_tissue_expression_multimapping(
testis_specific_in_multimapping_analysis$external_gene_name,
multimapping = FALSE, units = "log_TPM")
normal_tissue_expression_multimapping(
testis_specific_in_multimapping_analysis$external_gene_name,
multimapping = TRUE, units = "log_TPM")
5.1.3 testis_expression()
Allows to visualise gene expression in all different testis cell type, somatic or germ, using data from the adult human testis transcriptional cell atlas.
Using the testis cell type defined for each CT genes, we visualise clearly that genes mainly expressed in an early stage of spermatogenesis aren’t expressed later and vice-versa.
early_CT <-
dplyr::filter(CT_genes, testis_cell_type %in%
c("SSC", "Spermatogonia", "Early_spermatocyte"))
testis_expression(early_CT$external_gene_name,
cells = "germ_cells")## `use_raster` is automatically set to TRUE for a matrix with more than
## 2000 columns You can control `use_raster` argument by explicitly
## setting TRUE/FALSE to it.
##
## Set `ht_opt$message = FALSE` to turn off this message.
late_CT <-
dplyr::filter(CT_genes, testis_cell_type %in%
c("Late_spermatocyte", "Round_spermatocyte",
"Sperm1", "Sperm2"))
testis_expression(late_CT$external_gene_name,
cells = "germ_cells")## `use_raster` is automatically set to TRUE for a matrix with more than
## 2000 columns You can control `use_raster` argument by explicitly
## setting TRUE/FALSE to it.
##
## Set `ht_opt$message = FALSE` to turn off this message.
5.1.4 HPA_cell_type_expression()
Allows to visualise gene expression in all different healthy cell type, comparing different somatic cells to germ cells, using data from the human proteome atlas.
Visualising all CT_genes, the specificity to germ cells only is quite clear.
HPA_cell_type_expression(units = "scaled")## see ?CTdata and browseVignettes('CTdata') for documentation## loading from cache## Warning: 48 out of 298 names invalid: KAZN-AS1, NFIA-AS1, HHLA3-AS1, SPATA42,
## LINC01356, LINC01940, NECTIN3-AS1, AADACL2-AS1, LINC01206, LINC01096,
## LINC02475, LINC02377, LINC01095, LINC01511, LINC02241, LINC02228,
## LINC02105, LINC01947, SLIT3-AS1, MIR548A1HG, LINC01016, TDRG1,
## LINC03004, SEC61G-DT, GTF2I-AS1, NRG1-IT1, IKBKB-DT, LINC01030,
## LINC00200, LINC01518, MRGPRG-AS1, TMEM132D-AS1, FAM230C, LINC00408,
## LINC01467, LINC01193, LINC01413, CT62, LINC02169, FLJ36000, LINC02074,
## LINC01970, LINC00470, LINC00668, LINC02864, LINC01548, DSCR8, FAM230I.
## See the manual page for valid types.
5.2 Expression in cancer cells and samples
5.2.1 CCLE_expression()
Allows to visualise gene expression in different histological types of CCLE cancer cell lines. We can thus compare genes that are or not activated in tumoral cell lines.
- Applied to CT genes frequently activated in CCLE cell lines, more than 5% of CCLE cell_lines are expressing each gene. A string signal is visible.
frequently_activated <- dplyr::filter(
CT_genes,
percent_of_positive_CCLE_cell_lines >= 5)
CCLE_expression(
genes = frequently_activated$external_gene_name,
type = c(
"lung", "skin", "bile_duct", "bladder", "colorectal",
"lymphoma", "uterine", "myeloma", "kidney",
"pancreatic", "brain", "gastric", "breast", "bone",
"head_and_neck", "ovarian", "sarcoma", "leukemia",
"esophageal", "neuroblastoma"),
units = "log_TPM")
- Applied to CT genes not frequently activated in CCLE cell lines, more than 95% of cell lines are not expressing each genes. The lack of expression is very clear.
not_frequently_activated <- dplyr::filter(
CT_genes,
percent_of_negative_CCLE_cell_lines >= 95)
CCLE_expression(
genes = not_frequently_activated$external_gene_name,
type = c(
"lung", "skin", "bile_duct", "bladder", "colorectal",
"lymphoma", "uterine", "myeloma", "kidney",
"pancreatic", "brain", "gastric", "breast", "bone",
"head_and_neck", "ovarian", "sarcoma", "leukemia",
"esophageal", "neuroblastoma"),
units = "log_TPM")
5.2.2 CT_correlated_genes()
A function that use expression data from all CCLE cell lines and returns genes correlated (or anti-correlated) with specified CT gene.
Here with MAGEA3, we can see all genes (CT or not) whose expression is correlated with it, like MAGEA6.
CT_correlated_genes("MAGEA3", 0.3)## Warning: Removed 4 rows containing missing values or values outside the scale range
## (`geom_point()`).## Warning: ggrepel: 157 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
5.2.3 TCGA_expression()
Allows to visualise gene expression in cancer samples from TCGA. We can thus for example compare with activation in CCLE cell lines.
- Applied to CT genes frequently activated in CCLE cell lines
TCGA_expression(
genes = frequently_activated$external_gene_name,
tumor = "all",
units = "log_TPM")## `use_raster` is automatically set to TRUE for a matrix with more than
## 2000 columns You can control `use_raster` argument by explicitly
## setting TRUE/FALSE to it.
##
## Set `ht_opt$message = FALSE` to turn off this message.
- Applied to CT genes not frequently activated in CCLE cell lines
TCGA_expression(
genes = not_frequently_activated$external_gene_name,
tumor = "all",
units = "log_TPM")## `use_raster` is automatically set to TRUE for a matrix with more than
## 2000 columns You can control `use_raster` argument by explicitly
## setting TRUE/FALSE to it.
##
## Set `ht_opt$message = FALSE` to turn off this message.
When visualising only one tumor, there is a separation between tumoral and peritumoral tissue.
TCGA_expression(
genes = frequently_activated$external_gene_name,
tumor = "LUAD",
units = "log_TPM")
5.3 Methylation analysis
5.3.1 DAC_induction()
Allows to visualise gene induction upon DAC treatment in a series of cell lines. We can see the difference between CT genes regulated by methylation or not
- Applied to genes controlled by methylation
controlled_by_methylation <- dplyr::filter(CT_genes, regulated_by_methylation)
DAC_induction(genes = controlled_by_methylation$external_gene_name)
- Applied to genes not controlled by methylation
not_controlled_by_methylation <- dplyr::filter(
CT_genes,
!regulated_by_methylation)
DAC_induction(genes = not_controlled_by_methylation$external_gene_name)
5.3.2 normal_tissues_methylation()
Gives the methylation level of all CpGs located in each promoter region (by default defined as 1000 nt upstream TSS and 200 nt downstream TSS) in a set of normal tissues. Can be used to analyse more precisely the methylation of a particular promoter in normal tissues as methylation values are given CpG by CpG.
- Applied to a gene controlled by methylation
normal_tissues_methylation("MAGEB2")
- Applied to a gene not controlled by methylation
normal_tissues_methylation("LIN28A")
5.3.3 normal_tissues_mean_methylation()
Gives the mean methylation level of CpGs located in each promoter region (by default defined as 1000 nt upstream TSS and 200 nt downstream TSS) in a set of normal tissues. When comparing genes controlled or not by methylation, we can see a specific methylation pattern in the first group.
- Applied to genes controlled by methylation
normal_tissues_mean_methylation(
genes = controlled_by_methylation$external_gene_name)
- Applied to genes not controlled by methylation
normal_tissues_mean_methylation(
genes =
not_controlled_by_methylation$external_gene_name)
5.3.4 TCGA_methylation_expression_correlation()
Shows the correlation between gene expression and promoter methylation in TCGA samples.
- Applied to a gene controlled by methylation
TCGA_methylation_expression_correlation(
tumor = "all",
gene = "TDRD1")
- Applied to a gene not controlled by methylation
TCGA_methylation_expression_correlation(
tumor = "all",
gene = "LIN28A")
6 Bibliography
Aguet, F., Anand, S., Ardlie, K. G., Gabriel, S., Getz, G. A., Graubert, A., Hadley, K., Handsaker, R. E., Huang, K. H., Kashin, S., Li, X., MacArthur, D. G., Meier, S. R., Nedzel, J. L., Nguyen, D. T., Segrè, A. V., Todres, E., Balliu, B., Barbeira, A. N., … Volpi, S. (2020). The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. https://science.sciencemag.org/content/369/6509/1318.abstract
Almeida, L. G., Sakabe, N. J., deOliveira, A. R., Silva, M. C. C., Mundstein, A. S., Cohen, T., Chen, Y.-T., Chua, R., Gurung, S., Gnjatic, S., Jungbluth, A. A., Caballero, O. L., Bairoch, A., Kiesler, E., White, S. L., Simpson, A. J. G., Old, L. J., Camargo, A. A., & Vasconcelos, A. T. R. (2009). CTdatabase: a knowledge-base of high-throughput and curated data on cancer-testis antigens. Nucleic Acids Research, 37(Database issue), D816–D819.
Barretina, J., Caponigro, G., Stransky, N., Venkatesan, K., Margolin, A. A., Kim, S., Wilson, C. J., Lehár, J., Kryukov, G. V., Sonkin, D., Reddy, A., Liu, M., Murray, L., Berger, M. F., Monahan, J. E., Morais, P., Meltzer, J., Korejwa, A., Jané-Valbuena, J., … Garraway, L. A. (2012). The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature, 483(7391), 603–607.
Cancer Genome Atlas Research Network, Weinstein, J. N., Collisson, E. A., Mills, G. B., Shaw, K. R. M., Ozenberger, B. A., Ellrott, K., Shmulevich, I., Sander, C., & Stuart, J. M. (2013). The Cancer Genome Atlas Pan-Cancer analysis project. Nature Genetics, 45(10), 1113–1120.
Carter, J. A., Matta, B., Battaglia, J., Somerville, C., Harris, B. D., LaPan, M., Atwal, G. S., & Barnes, B. J. (2023). Identification of pan-cancer/testis genes and validation of therapeutic targeting in triple-negative breast cancer: Lin28a- and Siglece-based vaccination induces anti-tumor immunity and inhibits metastasis. In bioRxiv (p. 2023.05.09.539617).
Guo, J. et al. The adult human testis transcriptional cell atlas. Cell Res. 28, 1141–1157 (2018).
Uhlén, M. et al. Proteomics. Tissue-based map of the human proteome. Science 347, 1260419 (2015).
Thorvaldsdóttir, H., Robinson, J. T., & Mesirov, J. P. (2013). Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Briefings in Bioinformatics, 14(2), 178–192.
Wang, C., Gu, Y., Zhang, K., Xie, K., Zhu, M., Dai, N., Jiang, Y., Guo, X., Liu, M., Dai, J., Wu, L., Jin, G., Ma, H., Jiang, T., Yin, R., Xia, Y., Liu, L., Wang, S., Shen, B., … Hu, Z. (2016). Systematic identification of genes with a cancer-testis expression pattern in 19 cancer types. Nature Communications, 7, 10499.
Session information
## R version 4.4.0 RC (2024-04-16 r86468)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 22.04.4 LTS
##
## Matrix products: default
## BLAS: /home/biocbuild/bbs-3.20-bioc/R/lib/libRblas.so
## LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.10.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_GB LC_COLLATE=C
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## time zone: America/New_York
## tzcode source: system (glibc)
##
## attached base packages:
## [1] stats4 stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] SingleCellExperiment_1.27.0 SummarizedExperiment_1.35.0
## [3] Biobase_2.65.0 GenomicRanges_1.57.0
## [5] GenomeInfoDb_1.41.0 IRanges_2.39.0
## [7] S4Vectors_0.43.0 BiocGenerics_0.51.0
## [9] MatrixGenerics_1.17.0 matrixStats_1.3.0
## [11] CTexploreR_1.1.0 CTdata_1.5.0
## [13] BiocStyle_2.33.0
##
## loaded via a namespace (and not attached):
## [1] DBI_1.2.2 rlang_1.1.3 magrittr_2.0.3
## [4] clue_0.3-65 GetoptLong_1.0.5 compiler_4.4.0
## [7] RSQLite_2.3.6 png_0.1-8 vctrs_0.6.5
## [10] pkgconfig_2.0.3 shape_1.4.6.1 crayon_1.5.2
## [13] fastmap_1.1.1 magick_2.8.3 dbplyr_2.5.0
## [16] XVector_0.45.0 labeling_0.4.3 utf8_1.2.4
## [19] rmarkdown_2.26 UCSC.utils_1.1.0 tinytex_0.50
## [22] purrr_1.0.2 bit_4.0.5 xfun_0.43
## [25] zlibbioc_1.51.0 cachem_1.0.8 jsonlite_1.8.8
## [28] blob_1.2.4 highr_0.10 DelayedArray_0.31.0
## [31] parallel_4.4.0 cluster_2.1.6 R6_2.5.1
## [34] bslib_0.7.0 RColorBrewer_1.1-3 jquerylib_0.1.4
## [37] Rcpp_1.0.12 bookdown_0.39 iterators_1.0.14
## [40] knitr_1.46 Matrix_1.7-0 tidyselect_1.2.1
## [43] abind_1.4-5 yaml_2.3.8 doParallel_1.0.17
## [46] codetools_0.2-20 curl_5.2.1 lattice_0.22-6
## [49] tibble_3.2.1 withr_3.0.0 KEGGREST_1.45.0
## [52] evaluate_0.23 BiocFileCache_2.13.0 circlize_0.4.16
## [55] ExperimentHub_2.13.0 Biostrings_2.73.0 pillar_1.9.0
## [58] BiocManager_1.30.22 filelock_1.0.3 foreach_1.5.2
## [61] generics_0.1.3 BiocVersion_3.20.0 ggplot2_3.5.1
## [64] munsell_0.5.1 scales_1.3.0 glue_1.7.0
## [67] tools_4.4.0 AnnotationHub_3.13.0 Cairo_1.6-2
## [70] grid_4.4.0 tidyr_1.3.1 AnnotationDbi_1.67.0
## [73] colorspace_2.1-0 GenomeInfoDbData_1.2.12 cli_3.6.2
## [76] rappdirs_0.3.3 fansi_1.0.6 S4Arrays_1.5.0
## [79] ComplexHeatmap_2.21.0 dplyr_1.1.4 gtable_0.3.5
## [82] sass_0.4.9 digest_0.6.35 SparseArray_1.5.0
## [85] ggrepel_0.9.5 farver_2.1.1 rjson_0.2.21
## [88] memoise_2.0.1 htmltools_0.5.8.1 lifecycle_1.0.4
## [91] httr_1.4.7 GlobalOptions_0.1.2 mime_0.12
## [94] bit64_4.0.5