| Objective:Screening and identifying the key genes of colorectal cancer,and studying the prognostic value and mechanism of the key gene PHLPP2.Methods:1)Sequencing data of all colorectal cancer patients were obtained from TCGA database,and paired data with cancer and adjacent tissues were sorted out by analyzing sample coding rules.Lima and edgeR were used to analyze the differential expression of the data,and the intersection of the results of Limma and edgeR was taken as the differentially expressed genes that related to colorectal cancer.The chromosome distribution of differentially expressed genes was analyzed by circlize,and the heterogeneity of expression of differentially expressed genes among individuals was analyzed by thermogram and cluster analysis.Based on the above differentially expressed genes,the protein-protein interaction network was constructed,and key genes were screened based on the characteristics of network nodes.RT-qPCR was used to verify the expression of key genes in colorectal cancer and adjacent tissues.The relationship between key genes and clinicopathological changes and overall survival was analyzed.2)The correlation between transcription of key genes and protein abundance was analyzed by protein mass spectrometry in colorectal cancer.Immunohistochemistry was used to detect the expression of PHLPP2 in185 cases of human colorectal cancer and adjacent tissues.The expression of PHLPP2 was semi-quantified based on the expression range and color rendering degree of PHLPP2.The correlation between the expression of PHLPP2 and clinicopathological factors was analyzed.Univariate Cox analysis was used to analyze the relationship between PHLPP2,clinicopathological factors and overall survival,and multivariate Cox analysis was used to screen out factors related to prognosis.The ROC curve was used to compare the accuracy of prognostic assessment between PHLPP2 expression and pathological factors.Nomogram was used to construct an individualized assessment tool for patients with colorectal cancer.The ROC curve was used to compare the accuracy of Nomogram with clinicopathological features,PHLPP2 expression and integrated clinicopathological features in prognostic evaluation.3)Western blot was used to detect the expression of PHLPP2 in six colorectal cancer cell lines.Cell lines that stable over expressed or silenced PHLPP2 were constructed in HCT116 and HT29,respectively.The transfection rate of lentivirus was observed by fluorescence microscopy,and the differences of PHLPP2 in the levels of protein and transcription were detected by Western blot and RT-qPCR.High throughput sequencing was used to investigate the expression profiles of Cont group and over-expression PHLPP2 group(Lv-PHLPP2 group)HCT116 cell,and gene enrichment analysis was carried out.Flow cytometry was used to detect the levels of reactive oxygen species(ROS)and the proportion of CD133~+cells in the two kinds of cells.The difference of the number of stem cell microspheres in the two groups was observed by sphere-forming,the sensitivity of the two groups to 5-FU and Oxaliplatin was tested by drug toxicity test,and the volume difference of transplanted tumors formed by two groups of cells in nude mice was detected by tumorigenesis experiment.Western blot and RT-qPCR were used to detect and compare the expression difference of CD44,CD133,Nrf2 and EPCAM in the two groups of cells.Lv-PHLPP2 HCT116 was disturbed with DMSO or Nrf2agonist Cyclopamine,respectively.Then,western blot was used to detect the expression of Nrf2 and stem cell related markers included CD44,CD133,EPCAM.Results:1)548 patients and 598 expression profiles were screened from 645patients with colorectal cancer in TCGA.Among them,50 patients have cancer-adjacent paired expression profiles,and 498 patients only have cancer tissue expression profiles.Limma and edgeR statistics showed a total of 1335 cross-genes,of which 447 were up-regulated and 888 were down-regulated.Differentially expressed genes were distributed evenly on chromosomes 1 to 22 and X.No differentially expressed genes were found that are come from Y chromosome.The expression of these genes showed good consistency among paired tissues,but there were significant differences in the expression of these genes in unpaired tumors.Four core subnets were found in the interaction network of differentially expressed genes.From the four subsets,the gene with the highest MCODE coefficient were extracted as the key genes:PHLPP2,ELF5,TIMP1 and CXCL3.The expression of these four genes was significantly correlated with the clinicopathological features and prognosis of the patients.2)The expression abundance of PHLPP2(R=0.31,p<0.0001)increased with the increase of its transcription of mRNA,the changing trend with a significant positive correlation.Immunohistochemistry showed that PHLPP2 protein was mainly expressed in the cell envelope and aggregated on the inner surface of the envelope.The expression of PHLPP2in colorectal cancer tissues was significantly lower than that in adjacent normal tissues(p<0.001),and its expression abundance was closely related to age,lymph node metastasis,lymph node positive rate,differentiation degree of tissue cases,TNM stage(p>0.05),but not to the clinicopathological parameters such as gender,tumor location,local infiltration,distant metastasis,number of lymph nodes dissected(p>0.05).Kaplan-Meier survival analysis showed that PHLPP2 expression could significantly affect the overall survival of patients with colorectal cancer(log-rank p<0.001).Multivariate Cox risk model analysis suggested that low expression of PHLPP2 was an independent risk factor for poor prognosis in patients with colorectal cancer.PHLPP2 has the same prognostic accuracy as local lymphatic metastasis(N stage)(p<0.001).Nomogram,an individualized prognostic assessment tool,was constructed by integrating PHLPP2 with clinicopathological features.The area under Nomogram's ROC curve was 80.1,which was more accurate than the prognostic evaluation of combined clinicopathological factors(p<0.001),and was able to stratify the overall prognostic risk of patients more significantly.3)The protein abundance and transcription level of PHLPP2 in HT29and HCT116 were significantly lower than the other four colorectal cell lines.The expression of PHLPP2 was interfered by lentiviral transfection in HT29 and HCT116,and the fluorescence transfection rate was more than90%.The expression of PHLPP2 protein in Lv-shPHLPP2 group was significantly decreased,while the expression of PHLPP2 protein in Lv-PHLPP2 group was significantly increased.Changing the expression of PHLPP2 can affect about 8%of the known transcriptional genes,including53%of the RNA,27%of the pseudogenes,3%of the lncRNA and 1%of the microRNAs.It can also change the expression of antioxidant genes such as PRDX1,GCLC,GPX2 and GLCM.GSEA analysis showed that the genes affected by the expression of PHLPP2 were significantly enriched in the Nrf2 pathway.Compared with Cont group,ROS level in Lv-PHLPP2 group increased significantly,the proportion of CD133~+cells decreased significantly,the formation of microspheres decreased significantly,the sensitivity to 5-FU and Oxaliplatin increased significantly,the ability of metastasis,invasion and scratch healing decreased,and the volume of subcutaneous transplanted tumors in nude mice decreased significantly.Intervention of PHLPP2 can affect the expression of stemness markers and Nrf2 in colorectal cancer cells.Adding Sulforaphane to Lv-PHLPP2 group can partly reverse the decrease of Nrf2 caused by overexpression of PHLPP2,and increase the expression of CD44,CD133and EPCAM proteins.Conclusion:In colorectal cancer,accurate differential expression analysis and protein interaction network can be used to screen key genes more effectively.The key gene PHLPP2 is an independent prognostic marker of colorectal cancer.Integrating the expression of PHLPP2 can further improve the prognostic evaluation of clinicopathological features.In colorectal cancer,PHLPP2 may be involved in the regulation of"stemness"tumor cells through Nrf2,which leading to the changes malignant biological behavior such as invasion,metastasis and drug resistance. |
PDF Full Text Request