Integrated analysis of genomic and transcriptomic data in clear-cell renal cell carcinoma

摘要

Renal cell carcinoma (RCC) is the most widespread kidney tumor, which originates mostly from distal kidney tubules. RCC is the major mortality cause in excretory system cancer in adults, and constitutes about 80% of various kidney cancers. The 5-year survival rate is 60-70%, but lowers significantly in case of metastasis formation. The tumor is relatively resistant to chemotherapy and radiation therapy, but responds to immunotherapy. Target drugs (e.g. sunitinib, bevacizumab, @a-interferon, sorafenib) are more preferable in RCC therapy. In most cases RCC is caused by obesity, smoking, arterial hypertension and hereditary factors. 4 patients of Voronezh Regional Clinical Oncology Center, aged from 50 to 75years old were enrolled in this study. All were diagnosed with renal cell carcinoma confirmed by immunohistochemical analysis. Gene expression profiling was performed using Affymetrix GeneAtlas system with Affymetrix Hunam Gene 1.1 ST DNA- microarrays. Data moralization, statistical analysis and differentially expressed genes (DEG) list creation were performed using Partek Genomics Suite v. 6.6. DNA samples were sequenced on Ion Proton sequencer with Comprehensive Cancer Panel primer pool. Subsequent pathway analysis and biological interpretation were conducted using Ingenuity Pathway Analysis system and Ion Reporter Suite. 3528 genes were differentially expressed with expression change more than 2 times, and 351 genes changed their expression more than 5 times (p-value <0.05). Only mutations causing either amino-acid substitution in corresponding protein, open reading frame shift, or truncated transcript formation were taken into account. Targeted DNA sequencing revealed 99 common mutations in normal tissues of all test subjects. 14 mutations were localized in SDHB, TRIM33, PDE4DIP, PBX1, ABL2, MTR, VHL, ROS1, PRKDC, CSMD3, MLLT10, TRIP11, PER1, genes and were tumor-specific in all patients. 33 mutations were localized in promoter regions of EGFR, PDGFRA and HNF1A genes. The most representative metabolic and signaling pathways according to Ingenuity knowledge base were ''FXR/RXR Activation'', ''Atherosclerosis Signaling'', ''LXR/RXR Activation'', ''Production of NO and ROS in Macrophages'', ''Cell migration and adhesion''. Five most downregulated genes among all patients were CALB1 (-187), HPD (-133), KNG1 (-126), SLC36A2 (-126) and PAH (-122). Five most upregulated genes were TNFAIP6 (+34), ANGPT2 (+23), SERPINE1 (+22), CP (+20), HILPDA (+20). Genes with mutations and differentially expressed genes were simultaneously included in pathways analysis in order to generate gene networks with possible upstream regulators (mutated genes) included. This allowed observing a number of gene interactions not present in existing reports. This method of genomic and transcriptomic data integration allowed us to determine the sources of mRNA level variation. A number of mutations and specific DEGs discovered in this study are not registered in existing databases of annotated mutations, which allows us to propose population heterogeneity of RCC causes. These DNA mutations and mRNA level changes could be used as predictive biomarkers of renal cell carcinoma.