Construction And Verification Of A Multi-omics-based Prognostic Classifier For Clear Cell Renal Cell Carcinoma

Purpose:With the prolongation of life expectancy,malignant tumors have become one of the main causes of harm to human health,seriously threatening the health of Chinese people.Renal cell carcinoma(RCC)is one of the most common tumors in the urinary system.According to the Global Cancer Observatory,RCC has about 400,000 new cases and175,000 deaths each year.RCC can be roughly divided into clear cell renal cell carcinoma(cc RCC or KIRC),Kidney renal papillary cell carcinoma(KIRP)and Chromophobe renal cell carcinoma(KIRP)according to the pathological type.carcinoma,KICH),among which cc RCC is the most common pathological type in RCC,accounting for about 70-80% of the overall RCC.It is worth noting that the 5-year survival rate of patients with localized cc RCC is 65%,but once metastasis occurs,the survival rate of patients will drop to 10-20%.The tumor heterogeneity of cc RCC is the main factor leading to the poor prognosis of these patients.In recent years,based on the genomic and metabolomic data analysis of large cohorts of renal cancer,it was found that there are clonal evolution and metabolic reprogramming in cc RCC.It is worth noting that cc RCCs with similar tumor stages and grades in the clinic usually have very different prognosis,including drug resistance and metastasis,which brings great difficulties to clinical stratified diagnosis and treatment.How to identify high-risk groups in cc RCC patients and guide individualized diagnosis and treatment of patients;and how to identify effective prognostic markers are urgent issues to be solved in current cc RCC research.Methods:In this study,by collecting the multi-omics data of 258 cc RCC patients in the TCGA database,firstly,based on Cox stepwise regression calculation,the eigenvalues associated with the prognosis of cc RCC patients at each omics level were selected,and then the prognosis-related multi-omics data were classified based on 9 clustering algorithms.The omics data were integrated with cluster analysis to determine the final risk classification of cc RCC.The subtypes were compared in a multi-omics dimension to identify salient features of the malignant phenotype of cc RCC and explore potential therapeutic targets.At the same time,a prognostic risk calculation model was constructed based on the above characteristics to evaluate the survival rate of patients.In addition,this study conducted in vitro and in vivo experiments on SNRPA1,a key molecule in the malignant progression of cc RCC,and evaluated the biological function of SNRPA1 by Cas9 knockout system,and further confirmed the cancer-promoting effect of SNRPA1 in animal experiments.Results:Based on multiple algorithms and multi-dimensional omics data,we divided cc RCC patients into two subtypes,CS1 and CS2.The prognosis of patients with CS1 subtype was significantly worse than that of CS2 group,and there was an evolutionary relationship between CS2 and CS1.These progression events include biological processes such as cytokeratinization,epithelial cell differentiation,extracellular matrix remodeling,and suppression of secretory phenotype;among them,the mutation frequencies of VHL,PBRM1,and ARID1 A in CS1 are significantly reduced in CS1 subsets,and these genes are frequently mutated.is a driver of the malignant phenotype of cc RCC.At the same time,there are also significant differences in the immune infiltration profiles of the two subgroups.The degree of immune infiltration in the CS1 subgroup is generally better than that in the CS2 subgroup,and a variety of immune pathways in this subgroup are activated,suggesting that the immune microenvironment is present in cc RCC.Negative regulation;and CS1 is more resistant to immunotherapy and targeted therapy.Finally,the risk scores constructed based on subtypes can achieve higher predictive performance in the training set and validation set.At the same time,we systematically studied the specific marker SNRPA1 in the CS1 subgroup,and found that this gene is highly expressed in cc RCC tumor tissues,which can promote the invasion,metastasis and target drug resistance of cc RCC cell lines.The efficacy of targeted and immunotherapy in cc RCC patients,knocking out this gene can reverse the malignancy of tumor cells and reduce the volume of subcutaneous tumors.Conclusion:This study first identified a risk classifier based on multi-omics data based on multiple machine learning algorithms,which can guide the risk stratification and medication selection of cc RCC patients.At the same time,the analysis of SNRPA1,a key factor of high-risk renal cancer subtype CS1,showed that this molecule can activate pathways such as Myc and E2 F to promote the malignant progression of renal cancer cells and affect the prognosis of patients.SNRPA1 is expected to become a new diagnosis and treatment target for cc RCC.