Driver genes associated to Broad Copy Number Aberrations in solid cancers

Large numerical and structural chromosomal aberrations (aneuploidy) are intrinsic features of tumour cells and cell transformation. The chromosomal aberrations are collectively called Broad Copy Number Aberrations (BCNAs) and mainly include aneuploidy (i.e. an incorrect number of one or more chromosomes like monosomy, trisomy, or tetrasomy), and gains or losses of chromosomal arm as well as of large chromosomal segments with a size > 50% chromosomal arm. Such somatic chromosomal abnormalities are very common in multiple cancers and have been suggested to represent a type of driver aberration. Consequently, the cancer phenotype is in part based on these changes and mainly on the changes in transcript levels correlated to gene dosage. It is known that differences in transcripts dosage are among the common causes of cancer progression, especially when somatic chromosomal aberrations occur into the cells. The molecular mechanism linking them to cancerogenesis and their cancer-driving properties are still largely unknown. To investigate the BCNAs in solid tumors such as in Breast cancer (BRCA), Colon Adenocarcinoma (COAD) and Glioblastoma multiforme (GBM) we performed a massive bioinformatic analysis based on the identification of the specific aneuploidy-profiling for each carcer type as well as the investigation of the gene-dosage effect due to the gains of whole chromosome arm. In breast cancer, the most recurrent chromosomal aberrations are derivative chromosome der(1;16), deleted 16q and isochromosome 1q. To investigate the role of these aneuploidies, we performed a comparative and integrative bioinformatics analysis involving cytogenetic SNP array data (1084 samples), RNA-seq data (1222 samples in total subdivided in 1072 primary tumor and 99 normal breast tissue) and single-point mutation (WES-seq) data from The Cancer Genome Atlas (TGCA). Breast cancer adenocarcinomas were classified in five groups, called 1,16 chromogroups, according to different patterns of arm-level aberrations of chromosome 1 and 16. Cancer samples were further classified in histological subtype (e.g. Ductal and Lobular) as well as in intrinsic molecular subtypes (e.g. Luminal A, Luminal B, Her2+, Normal-like and Basal-like). The study design allowed identification of genes whose transcription is sensitive to copy number abnormalities of chr1 and 16. The pathway enrichment analysis has highlighted dysregulated pathways involved in processes such as Notch-signaling, formation of the beta-catenin complex and Wnt-signaling. The main candidate driver genes are BCL9, PYGO2 (Wnt enhanceosome components), APH1A, PSEN2, NCSTN (γ-secretase subunits), and CDH1 (involved in cell adhesion). The bioinformatics results were verified on cell models bearing chr 1/16 aberration (CAL148) or not bearing it (CAL51). We performed siRNA interfering and pharmacological experiments targeting the γ-secretase complex. The results show that CAL148 are sensitive to PF3084014, a γ-secretase inhibitor with IC50 10uM. The potential impact of siRNA interfering is analyzed by qPCR, and cell viability MTT assay. PSEN2 and APH1A co-silencing produces a significant decrease of cell viability in CAL 148 cells. In Colon Adenocarcinoma the most recurrent chromosomal aberrations regard the gain of chromosome 7, 8, 13 and chromosome 20. To investigate the gene-dosage effect of the so called Over-UpT, we downloaded 1) RNA-seq data from The Cancer genome atlas project (TCGA) obtaining 480 COAD samples and 41 mucosal normal samples (n = 521) 2) SNP 6.0 arrays data obtening 439 samples from cBioPortal for cancer genomics. In order to assess the presence of known cancer genes among Over-UpT genes, we performed a comparison with three different cancer-related gene lists and a significant enrichment of cancer genes among Over-UpT was observed. In order to obtain more information on the relevance of Over-UpT in the CRC phenotype we evaluated the presence of aberrant enhancer activity, as indicated by enrichment of histone markers, and of CRC fitness genes, according to CRISPR/Cas9 functional screenings. Indeed, it has been suggested that malignant transformation of colon is accompanied by widespread locus-specific gains of enhancer activity, called variant enhancer loci (VELs) and a significant enrichment of recurrent gained VEL was detected among Over-UpT genes in specific chromosomes. Lastly, we tested whether individual genes associated with Over-UpT might represent cancer dependencies. We took advantage of publicly available data from a recent study identifying ‘fitness genes’ in CRC to show a significant enrichment of cancer fitness genes. These results support the hypothesis that the chromosomal density of overexpressed cancer fitness genes might play a significant role in the selection of gained chromosomes during cancer evolution. Analysis of functional pathways associated with OverT suggest that some multi-subunit protein complexes (eIF2, eIF3, CSTF and CPSF) are candidate targets for silencing transcriptional therapy. At the end, in Glioblastoma multiforme the most recurrent chromosome aberration regards the gain of chromosome 7 alone or in combination with focal amplification of EGFR locus (7p11.22). The aim of the project is to analyse the role of the aberrant chromosome 7 by focusing on the high transcripts levels due to the selective advantage in GBM to have three copies of chromosome 7 and then to discover which is the cluster of genes that undergoes the gene-dosage effect. A massive bioinformatics pipeline was performed by analyzing 1) 144 GBM samples (only primary tumour) and 5 normal brain tissue from TCGA (RNA-seq counts data) 2) 60 samples from Whole Exome Sequencing (WES) data and 3) 592 samples molecular cytogenetic data (SNP 6.0). 467 OverT genes are enriched in Chr7-gain/ 7p-FA+ group and, among these, 40 genes were located on chr7p11.2. It is relevant the involvement of biological functions in mitochondria based on genes cooperation chr19/chr7/7-FA+. The results about BRCA, COAD and GBM highlight the importance to investigate the aneuploidy-profiling in cancer. Future efforts will concern the application of analysis to new and more massive technologies such as single cell-RNAseq as well as the translation of bioinformatic goals to cell model to identify new therapeutic-drug targets.