| Background:Colorectal cancer(CRC)is one of the most common malignancies and its incidence among young people has been increasing every year.Recent investigations have shown a dramatic decrease in the patient’s 5-year survival rate following the progression from an early adenoma to a malignant CRC.As one of the most common histological types of CRC,colon adenocarcinoma(COAD)lesions are mostly benign in the early stages and are invasive and metastasized in further development,spreading to adjacent organs,leading to high mortality and recurrence rates.In terms of treatment,radical surgical resection combined with chemotherapy and radiotherapy is still the preferred treatment,but most patients cannot tolerate the side effects of chemotherapy,and to some extent,it will affect the physical condition of patients,leading to a further decline in the cure rate.However,there is still no reliable biomarker for prognosis evaluation,which remains a challenge for patients with COAD to choose an effective treatment.Therefore,it is urgent to establish brand-new and stable prognostic evaluation indicators in order to better predict the prognosis of patients with COAD and guide clinical treatment.RNA-binding proteins(RBPs),the protein that binds to RNA in eukaryotic cells,are key factors in the post-transcriptional modification process and play vital regulatory roles in some important cellular processes,such as RNA splicing,modification,transport,localization,stability,degradation and translation.Due to the all-round and multifunctional regulation of RBPs,even small changes can cause significant physiological and pathological effects.Therefore,the imbalance of RBPs is closely related to many human diseases,especially playing a key role in the occurrence and development of most malignant tumors.At present,there is still a lack of systematic research on the prognostic value and pathogenesis of RBPs in COAD.Therefore,systematic research on subtype classification,immunotherapy prediction,molecular biology level verification of core gene expression level and clinical prognostic value exploration of COAD patients will fill some gaps in the clinical and scientific research fields and bring good news to COAD patients.Objective:Based on bioinformatics analysis,the gene expression characteristics of COAD patients were described and the markers with obvious predictive value for COAD prognosis were screened.The prediction results were further verified by detecting the expression of key markers in tissues of clinical COAD patients and their paracancerous tissues with immunohistochemical experiments.Finally,the expression of the key markers in the COAD tumor cell line is verified by a molecular biology experiment,and further explore whether inhibition of its expression level can affect the migration and invasion ability of tumor cell lines and induce apoptosis of tumor cells.Materials and Methods:1 Data processing such as format unification,log conversion and batch correction were performed on the gene expression profiling data of COAD patients obtained from the TCGA database and the GEO database and relevant clinical information materials based on computer programming languages such as R language and Perl language.Through extensive early literature retrieval,RNA-binding protein-coding genes(RBPs)were collected and sequenced for subsequent research.2 The molecular subtype characteristics of different clinical samples were clustered based on non-negative matrix decomposition algorithm.The "survival" packet and "NMF" packet in R language were used to cluster COAD samples according to the expression spectrum of RBPs.Through the above computer language,survival analysis of molecular subtypes,analysis of differences in clinical characteristics,and immune correlation analysis(immune cell infiltration analysis and immune checkpoint analysis)were performed.3 To construct an RBP prognostic risk prediction model for COAD patients based on Cox proportional hazard regression model.The data were randomly divided into a training set and a verification set,and COAD patients were divided into high-risk group and low-risk group according to the median risk score in R language analysis.The training and validation data sets were subjected to survival analysis using the Kaplan-Meier method and the ROC curve was constructed to evaluate the predictive ability and accuracy of the prognostic risk model.The "time ROC" package was used to compare this model with a reported prognosis model.4 The expression of core markers in COAD tissue and its paracancerous tissues was detected based on the online databases of UALCAN and HPA as well as immunohistochemical experiments.And aims to verify the results of bioinformatics analysis from a clinical point of view.5 The differences of the transcription and translation levels of the predicted core genes in the colon adenocarcinoma cell line are verified based on the molecular biology experiments.We further explored its role in tumor cells by inhibiting the expression of core genes and using basic experimental techniques such as cell migration,invasion and apoptosis.Results:1 Through data collection and processing,expression data of 56753 genes in a TCGA-COAD cohort including 41 paracancerous samples and 473 tumor samples were obtained in this study.The GSE39582 cohort included expression data for 20174 genes from 19 paracancerous samples and 566 neoplastic samples.Based on the literature,we collected and sequenced 1542 recognized genes encoding RNA-binding proteins.2 Based on non-negative matrix factorization clustering algorithm,two molecular subtypes were obtained and their characteristics were described.Patients with subtype 1were more associated with longer survival.Analysis of clinical data revealed earlier tumor subtypes were included in subtype 1.Analysis of immune infiltration showed that subtype 1 had more CD8+T cells,CD4+T cells,NK cells,dendritic cells,and mast cells,and fewer γδT cells and M0 macrophages.In addition,all immune-related checkpoints had low expression in subtype 1.3 Based on the Cox proportional hazard regression model,we constructed a new RBPs-related prognostic model using the training set data,which contained a total of eight RBPs genes(including SRP14,DHX15,MRPS7,CLK1,LARS2,C11orf68,RRNAD1,and CWF19L2).This model can obviously distinguish high-risk group with poor prognosis from low-risk group,and this prediction potential is confirmed in the verification set data.In addition,in terms of immunoassay,patients in the low-risk group responded better to immunotherapy and patients in the high-risk group were more likely to undergo immune evasion.4 Protein levels of DHX15 in COAD tissue were confirmed to be higher than normal colonic tissue based on online databases UALCAN and HPA.In addition,immunohistochemical analysis of cancer and paracancerous tissue sections from 20 pairs of COAD patients collected from Department of Colorectal Surgery,The First Affiliated Hospital of Jinzhou Medical University.The results showed that the protein level of DHX15 in COAD tissues was significantly increased.5 The m RNA expression levels of the core gene DHX15 in DLD-1,SW480,SW48,HCT116,HT29(human colonic adenocarcinoma cells)and NCM460 cell lines(human normal intestinal epithelial cells)were determined by q RT-PCR.The results showed that compared with the NCM460 cell line(human normal intestinal epithelial cells),the m RNA and protein expression levels of DHX15 in the colon cancer cell line were significantly increased(P < 0.05).After the expression of DHX15 is inhibited,the migration,invasion and apoptosis of tumor cells are all changed to different degrees.Conclusions:1 In this study,we analyzed the differential changes in gene expression of patients with COAD when COAD occurred based on a variety of computer programming languages,performed clustering analysis on the molecular subtype characteristics of different clinical samples,delineated the gene expression spectrum among different subtypes,and constructed a brand new and stable RBPs gene-related prognosis evaluation model,which can be used to predict the prognosis of patients with COAD and the immunotherapy response.This model has great clinical potential in predicting the disease development trend of COAD.2 This study further verified the prediction results of bioinformatics through online database and immunohistochemical experiments,that is,DHX15 was significantly up-regulated in COAD patients,indicating that DHX15 is expected to be a biomarker for prognosis evaluation of COAD patients.3 The expression levels of DHX15 determined in this study based on thorough analysis were significantly up-regulated in the COAD cell line and,when knocked down,significantly affected the tumor cell phenotype(cell migration,invasion,and apoptosis).This provides new insights into the pathogenesis of COAD. |
PDF Full Text Request