| IntroductionColorectal cancer and metastatic colorectal cancer significantly reduce the quality of life of patients,seriously threaten the lives and health of patients,and bring a heavy financial burden to the families of patients.With the popularity of colonoscopy,patients with early detection of colorectal cancer generally have a good prognosis.However,about20% of colorectal cancer patients have already had tumor metastasis at the first diagnosis.For these patients,the 5-year survival rate was only 13%,and the location of metastasis and the number of metastatic lesions were closely related to the survival time of patients.The main target organs of colorectal cancer metastasis are liver and lung,and end stage colorectal patients may also have multiple systemic metastases including bone metastasis and brain metastasis.However,the underlying mechanisms that determine the pattern of colorectal cancer metastasis have not yet been elucidated.The liver is rich in blood flow,and the portal venous system is the main source of its blood flow.Clinical studies have shown that intestinal venous drainage into the portal venous system may be closely associated with liver metastasis of colorectal cancer.The lung is also an important target organ for many different types of primary tumor metastasis,partly because our entire cardiac output circulates through a network of pulmonary capillaries,facilitating tumor cell masses to embed in capillaries and eventually exude and colonize.However,the pattern of tumor metastasis and the selection of target organs cannot be fully explained by blood or lymph node drainage.As early as the 19 th century,some researchers have proposed tumor metastasis "seeds" and "soil" hypothesis: disseminated tumor cells are "seed",tumor cell engraftment of tissues and organs are the "soil",disseminated tumor cells(" seeds "),will colonize(" soil ")only with suitable to the growth of tissue microenvironment.In clinical practice,it has been found that digestive system tumors are more prone to liver and lung metastases,uveal melanoma is particularly prone to liver metastasis,sarcoma is more prone to lung metastasis,and prostate cancer is more prone to bone metastasis,but liver and lung metastasis are less likely to occur.This indicates that the selection of the target organ for metastasis by tumor cells is not random,and different primary tumors have different preferences for the selection of the target organ for metastasis.Tumor cells have specific phenotypic characteristics at each step,and genetic and epigenetic changes acquired by primary tumor cells during metastasis may contribute to tumor cell-host interaction.Therefore,studying the similarity and difference of gene expression between primary tumor and metastatic tumor will help to analyze the selection of target organs in the process of metastasis,and provide a beneficial reference for the prevention and improvement of the treatment of metastatic tumor.ObjectiveNowadays,a large number of studies mostly use molecular biology methods to analyze the specific changes of a gene and its products in the process of promoting or inhibiting the transfer.Few studies focus on the overall transcription profiles of such genes or their products in the process of metastasis.This study intends to analyze the high-through GEO gene array data set of colon cancer metastasis samples using rigorous bioinformatics and statistical methods on a macro-scale to identify key genes and key signaling pathways that are closely related to liver and lung metastasis of colorectal cancer cells,and to construct regression models that are capable of predicting metastasis of colorectal cancer to liver or to lung.Materials & MethodsIn this study,three GEO gene expression datasets of human colorectal cancer samples and clinical colorectal cancer samples from our hospital were used as the research objects.The original data set of the gene expression chip of all selected samples were downloaded by using File Zilla software.R and Rstudio were used as analysis and research platform.The affy package was used to read the expression values of all known genes in all samples in the original file of the gene chip.The imput package is used to fill in missing values using the nearest neighbor averaging algorithm.Batch effects between the three datasets were removed via empirical Bayes algorithm by using the sva package.Differentially Expressed Genes(DEGs)were calculated between each sample group by using the mainstream gene differential expression software package limma.The criteria for DEGs are: Adjusted p value(FDR)< 0.01 and | log2(Fold Change)| > 1.An online protein interaction analysis platform STRING was used to map out proteins’ networks of DEGs related to liver metastasis and lung metastasis.Using Cytoscape software for the visualization of protein-protein interaction network.Using the Clue GO plug-in of Cytoscape software to perform GO biological process annotation and KEGG pathway analysis on differentially expressed genes in liver and lung metastases.Best subset regression and backward stepwise regression were used to further screen differential genes for liver metastasis and lung metastasis.Paraffin sections of liver and lung metastasis samples of colorectal cancer from our hospital were used to extract total RNA.Gene expression values of these samples were obtained by q RT-PCR.The ROC curve was used as a validation tool,and gene expression data of clinical samples was used as a validation dataset to evaluate the efficacy of the self-built regression model in predicting liver cancer and lung metastasis samples from colon cancer.Results1.Based on the inclusion and exclusion criteria,three GEO datasets(GSE68468,GSE41258,and GSE49355)were selected for analysis.A total of 712 samples of gene expression array data were included,including 392 primary colon cancer tissues,127 normal colon tissues,113 liver metastatic tissues,26 normal liver tissues,40 lung metastatic tissues,and 14 normal lung tissues.Through group comparison,1832 DEGs of [liver metastasis tissue vs.Normal liver tissue] were found;1243 DEGs of [pulmonary metastasis tissue vs.Normal lung tissue] were found;1765 DEGs of [normal liver tissues vs.Normal colon tissue] were found;1242 DEGs in [normal lung tissue vs.Normal colon tissue] were found;680 DEGs in [liver metastasis tissue vs.Normal colon tissue] were found;602 DEGs in [lung metastasis Tissue vs.Normal colon tissue] were found;482 DEGs in [primary colorectal cancer tissue vs.Normal colon tissue] were found.2.Using the Venn diagram tool,104 DEGs related to colorectal cancer liver metastasis and 135 DEGs related to colorectal cancer lung metastasis were finally screened.Among them,45 DEGs were shared by liver metastasis and lung metastasis,59 DEGs were exclusively enriched in liver metastasis,and 90 DEGs were exclusively enriched in lung metastasis.There were significant differences in the DEGs profiles between liver metastasis and lung metastasis,suggesting significant differences in the mechanisms leading to liver metastasis and lung metastasis in colorectal cancer.3.Analysis of the KEGG pathway showed that the two KEGG pathways,including mineral absorption,protein digestion and absorption were mainly enriched in lung metastasis.KEGG pathways such as fat digestion and absorption,ascorbic acid and aldose metabolism,pentose and glucuronate interconversions were mainly enriched in liver metastasis.The IL-17 signaling pathway was enriched in both liver and lung metastases.4.GO biological process including regulation of c AMP-mediated signaling,response to zinc ion,detoxification of copper ion,cellular response to copper ion,cellular zinc ion homeostasis,regulation of cardiac muscle contraction by regulation of the release of sequestered calcium ion,positive regulation of calcium ion transmembrane transport and positive regulation of calcium ion transmembrane transporter activity were mainly enriched in lung metastasis;GO biological process including endodermal cell differentiation,cellular extravasation,artery morphogenesis,aorta morphogenesis and terms that concerning regulation of ion transmembrane transportation activity are shared in both liver and lung metastasis;no biological process is exclusively enriched in liver metastasis.5.The best subsets Regression of the 22 genes generated a 7 genes subset(including SPARC,COL1A2,MMP9,COL11A1,CXCL12,COL3A1 and THBS2)with the least genes number(7)and the highest adjusted R-squared value(0.63).The seven genes were subsequently applied to construct a logistic regression model and then assessed by Backward Stepwise Regression using step AIC,which reached an AIC(Akaike information criterion,the less the better)of 59.74.Cutting anyone belong of the 7 genes can’t further reducing AIC,which suggested the good fitness of the 7 genes regression model.6.48 clinical samples in the form of formalin fixed paraffin-embedded(FFPE)tissues were screened from archive storage of the pathology department in our hospital including19 cases of primary colon cancer,21 cases of liver metastasis,8 cases of lung metastasis.Total RNA was used to validate the expression level of the 7 discriminating genes via q RT-PCR.Then the expression value of the 7 discriminating genes were used as validation dataset to evaluate the performance and efficacy of the regression model.7.ROC curve was used to test the performance of the seven genes predicting regression model in predicting liver and lung metastasis of colon cancer via the expression data acquired by q RT-PCR of clinical FFPE samples.The results shown that the AUC is83.9%,which demonstrated the seven genes predicting regression model were capable of predicting the target organ of colon cancer metastasis with good efficacy.Conclusion1.Combined the KEGG pathway analysis and GO Biological Process analysis of all the shared and exclusive DEGs reveals that:(1)genes taken parts in metal ion migration and absorption,degradation of intercellular matrix are vital for lung metastasis;(2)genes participated in carbohydrate,protein,amino acid,lipid and energy metabolism are essential and important in forming liver metastasis;(3)genes taken parts in artery morphogenesis,forming tumor microenvironment by chemotactic factors,endodermal cell differentiation are the key genes for both liver and lung metastasis.2.The regression model established by the expression values of SPARC,COL1A2,MMP9,COL11A1,CXCL12,COL3A1 and THBS2 was capable of predicting the liver and lung metastasis of colorectal cancer with good efficacy.Research MeaningIn summary,we identified seven key genes and several signaling pathways in predicting liver/lung metastasis of colorectal cancer,which might make clues for exploring the mechanism that decide the target organ selection during metastasis and inspire new targets for metastasis inhibiting therapy.Introduction H.pylori infection can induce a significant inflammatory response in the gastric mucosa,accompanied by infiltration of a variety of immune cells including macrophages,lymphocytes,plasma cells and polynuclear leukocytes,but it cannot eradicate H.pylori,leading to chronic infection and sustained damage to gastric mucosal tissues.Although the mechanism concerning the chronic infection and persistent colonization of H.pylori in the gastric mucosa is not clear,existing studies suggest that the interaction of gastric epithelial cells and gastric mucosal immunity in H.pylori infection is a key contributing factor.Therefore,the analysis of the interaction between key molecules in gastric mucosal immunity and H.pylori and its mechanism is expected to provide useful references for the eradication of H.pylori,the treatment of chronic gastritis and the prevention of gastric cancer.Heparan sulfate proteoglycans(HSPGs)are not only an important component in the ECM,but also exist on the surface of almost all cells.The side chain of HSPGs provides numerous anchor sites for a variety of bioactive molecules(cytokines,chemokines,growth factors,enzymes,protease inhibitors,etc.).In the human and mouse,heparanase is the only enzyme that degrades HSPGs.Heparanase disconnects the side chain of HSPGs,releasing the bioactive molecules and ligand molecules bound to the side chain,and converting them into the active form.Heparanase is involved not only in cancer progression,metastasis,and angiogenesis,but also in a variety of inflammatory diseases.In the inflammatory response,heparanase regulate the response of inflammatory cells to inflammatory stimuli by releasing chemokines and cytokines that bind to HSPGs.Existing studies have found that heparanase are highly expressed in a variety of inflammatory diseases,but have different roles in different tissues and disease models.For example,heparanase plays a pro-inflammatory role in acute pancreatitis,acute vasculitis,acute glomerulonephritis and allergic pneumonia.Heparanase plays an anti-inflammation role in Alzheimer’s disease by inhibiting the inflammatory response and reducing macrophage clearance of amyloid beta.However,the interaction between heparanase and H.pylori infection in chronic gastritis remains unclear.Objective 1)To study the expression and source of heparanase in gastric mucosal tissue infected by H.pylori.2)To study the effect of heparanase on the severity of gastritis and the colonization of H.pylori.3)To study the effect of heparanase on the accumulation and function of infiltrating immune cells in chronic gastritis infected with H.pylori and its mechanism.Methods 1)The expression of heparanase in chronic H.pylori infection gastritis was detected by immunofluorescence and immunohistochemistry using fresh gastric mucosal tissues from patients with chronic H.pylori infection gastritis,chronic atrophic gastritis and healthy people.2)Heparanase knockout mice were obtained from Israel Institute of Technology.H.pylori PMSS1 strain was obtained from the First Affiliated Hospital of Nanchang University.Chronic gastritis model of H.pylori infection was established by using H.pylori PMSS1 strain.3)The colonization of H.pylori in gastric mucosa was identified by HE staining and Warthin-Starry silver staining.4)Immunofluorescence was used to detect the expression of pan-cytokeratin(the marker gene of epithelial cells)and heparanase,to reveal the cell source of heparanase in gastric mucosal tissues.5)Human gastric epithelial cell lines GES1 were used to detect the effect of H.pylori on the expression of heparin in gastric epithelial cells at different time points and different MOI by q RT-PCR and Western Blot.6)Heparanase knockout mice and wild-type mice were used to detect the effects of heparanase expression on gastric mucosal inflammation,immune cell infiltration,and pro-inflammatory cytokines after mouse infected with H.pylori for 8 weeks.7)Heparanase knockout mice,wild-type mice,and human chronic gastritis mucosal tissues was used to detect the influence of heparanase expression level on H.pylori colonization and the relationship between heparanase expression level and H.pylori colonization by Taqman probe PCR.8)q RT-PCR was used to detect the expression of surface markers in common inflammatory cells including NK cells(NK1.1,Gran B),dendritic cells(Langarin),macrophages(F4/80),and neutrophils(Ly6g)in gastric mucosa.Which revealed that the expression level of heparanase could affect the number of macrophages in the gastric mucosa of H.pylori-infected gastritis.9)By using flow cytometry,it was confirmed that heparanase deficiency can reduce the number of macrophages in the gastric mucosal tissue of H.pylori-infected gastritis.10)q RT-PCR was used to detect the effect of different heparanase expression levels on the expression of i NOS,markers of macrophage M1 polarization,and pro-inflammatory cytokines.Western Blot was used to detect the effect of different heparanase expression levels on the expression of i NOS protein.11)Western Blot was used to detect which signaling pathways would be activated by H.pylori stimulation to peritoneal macrophages at different heparanase expression levels.Results 1)Heparanase is highly expressed in human and mouse H.pylori infected gastritis tissues 2)In human and mouse H.pylori infected gastritis,the highly expressed heparanase is mainly derived from gastric mucosal epithelial cells 3)Heparanase can aggravate the inflammation of H.pylori induced chronic gastritis and promote H.pylori colonization 4)Heparanase promotes the infiltration of macrophages in chronic gastritis induced by H.pylori 5)Heparanase promotes macrophage polarization induced by H.pylori 6)Heparanase promotes H.pylori induced macrophage polarization by regulating the P38 MAPK and NF-κB pathwaysConclusion In our study,we revealed that H.pylori can promote the high expression of heparanase in gastritis,and the high expression of heparanase can promote the polarization of Macrophage M1 induced by H.pylori by regulating the P38 MAPK and NF-κB pathways,and increase the secretion of pro-inflammatory cytokines,thus aggregate gastritis. |
PDF Full Text Request