Molecular Characterization Of Papillary Craniopharyngiomas Based On Multi-omics

Purpose: Craniopharyngiomas(CPs)are rare,epithelial and histologically benign brain tumors that is most common in the sellar and suprasellar,accounting for 5-11% of the total intracranial tumors.According to histomorphology,it can be classified into two different subtypes,adamantinomatous craniopharyngioma(ACP)and papillary craniopharyngioma(PCP).PCP is highly prevalent in adults.Currently,the understanding of the molecular characteristics of PCP is not sufficient and comprehensive,especially the research on the tumor microenvironment of PCP is very few.Therefore,this study aims to analyze the molecular characteristics of PCP through multi-omics data and explore the role of tumor microenvironment in PCP,which may provide new ideas for clinical diagnosis and treatment.Methods: In this study,DNA methylation data and some RNA sequencing data were downloaded from the public database,and 18 paired tumor and blood samples of PCP patients were collected for whole-exome next generation sequencing,of which 9samples were simultaneously used for RNA sequencing.The best practice process recommended by Broad Institute was used to analyze somatic mutation,structural variation and copy number variation of PCP samples.Based on DNA methylation data,hierarchical clustering and principal component analysis(PCA)were used for comparison between PCP and normal pituitary samples,and Wilcoxon rank sum test was adopted for differential methylation analysis.The edge R algorithm was utilized to identify the differentially expressed genes(DEGs)in RNA sequencing samples among different brain tumors,and then GO/KEGG pathway enrichment analysis was performed for the DEGs.The hypomethylation-high expression genes were selected for PPI network analysis.CIBERSORT and x Cell algorithm were used to study the relative abundance of specific immune infiltrating cells in PCP samples.The single sample gene set enrichment analysis(ss GSEA)was employed to quantify the relative levels of signaling pathway activation.Results: We firstly analyzed the whole-exome sequencing data of 18 PCPs from scratch.It was showed that all the samples were CTNNB1 wild type,and 17 of them(94%)had BRAF mutations.The only one BRAF wild type sample was positive by VE1 immunohistochemistry.Heterogeneity of PCP samples was presented in structural variation analysis.There was no significant difference in the mutant profiles between children and adults,both of which involved MAPK signaling pathways.Secondly,we analyzed the DNA methylation data based on hierarchical clustering and PCA,and found that the DNA methylation patterns of PCP and normal pituitary were significantly different.Compared with pituitary,the hypomethylated genes in PCP are significantly enriched in inflammation related pathways,including MAPK signaling pathway,TNF signaling pathway,NF-kappa B signaling pathway,etc.Thirdly,according to the analysis of RNA sequencing data,the differentially up-regulated genes in PCP compared with normal pituitary have been demonstrated that they were also enriched in inflammation related pathways.Furthermore,we found that hypomethylation-high expression genes in PCP were also enriched in inflammation related pathways.Next,PPI network analysis was performed on these hypomethylation-high expression genes,and 11 key genes related to inflammation were finally obtained,including EGFR,SMAD3,JUN,MDM2,GSK3 B,RAC1,LYN,CDKN1 A,PRKCD,SP1 and MYC.Additionally,PCP,LGG and GBM were compared respectively because they all had BRAF V600 E mutation,and found they have different degree of inflammation.Finally,based on the integrated analysis of whole-exome and RNA sequencing data,the relationship between the inflammatory degree of three brain tumors was further confirmed.Conclusions: In this study,the analysis results of whole-exon sequencing data suggested that PCP is an inflammation-related tumor with similar pathogenesis in children and adults.Analysis of DNA methylation and transcriptome sequencing data provided further evidence that PCP is a tumor associated with inflammation.Through comparison with different brain tumors,the relationship of inflammation degree in some brain tumors have been proved that was: PCP > GBM > LGG.In a word,we hope our findings can be beneficial to the understanding of PCP and provide new ideas for the treatment of PCP.