| Background: The occurrence of gastric cancer(GC)is a multi-stage progression process,from chronic non-atrogastritis(non-CAG),chronic atrophic gastritis(CAG),intestinal metaplasia(IM),and gastric dysplasia(GD)gradually developing into cancer.The prognosis of early gastric cancer(EGC)is significantly better than that of advanced gastric cancer(AGC).Thus,it’s of vital importance to detect and treat GC in early or even pre-cancerous stages.According to Lauren’s histological classification,EGC can be divided into intestinal gastric cancer(IGC)and diffuse gastric cancer(DGC).For different stages of GC progression,the differences in histological classifications and disease locations may lead to genetic heterogeneity of the disease,which complicate our understanding of the occurrence and development of EGC and the screening of EGC pre-warning biomarkers.Spatial transcriptomic sequencing(ST-seq),an emerging technology based on a combination of sequencing and imaging technology,was able to detect most genes within the entire range of a tissue microarray,which could greatly improve our understanding of the pathogenesis of diseases.ST-seq had been applied in breast cancer,pancreatic ductal adenocarcinoma,liver cancer,prostate cancer and squamous cell carcinoma of the skin.However,the application of ST-seq in gastric cancer was rarely reported.The aim of our study was to explore the genetic heterogeneity and internal relationships in different stages of GC development,and construct a complete spatial expression profile of gastric diseases,providing clues for understanding the occurrence and development of EGC.OLFM4 was an extracellular matrix glycoprotein that includes a Olfactomedin domain with diverse N-terminal sequence and a 250-amino acid C-terminal sequence.OLFM4 was commonly expressed in the prostate,bone marrow,small intestine,pancreas and other organs.However,studies have found that OLFM4 was also highly expressed in a variety of cancers,including gastrointestinal cancer,squamous cell carcinoma of the head and neck,cervical neoplasia,and non-small cell lung cancer,and the high expression of OLFM4 was often negatively correlated with invasion and metastasis.Currently,there have been some studies on the correlation between OLFM4 and the differentiation degrees and clinical stages of GC,but few studies comprehensively analyzed the relationship between OLFM4 and clinical parameters of GC.In the present study,ST-seq was used to analyze the gastric mucosal samples from 8EGC patients with cancer and para-carcinoma normal tissue/IM tissues,constructing a complete ST-seq map of different stages of EGC progression,and exhibiting gene expression patterns at different stages.A series of in vitro and in vivo experiments were conducted to explore the function and possible mechanism of OLFM4 gene in GC.Objectives: We constructed the ST-seq map of different stages of GC progression to clarify the expression of hub gene in different gastric diseases.We demonstrated the correlations of OLFM4 and clinicopathological parameters of GC,and elucidated the influence and mechanisms of OLFM4 on biological behavior of GC cells.Methods:Part Ⅰ.Construction of spatial transcriptome maps at different stages of EGC1.The formalin fixed paraffin-embedded(FFPE)GC tissue collection: FFPE samples were collected from 8 patients with EGC after surgery.The eight cases were derived from different disease sites(i.e.,gastric body or antrum)and different histological classification(IGC or DGC).The patients all underwent endoscopic submucosal dissection(ESD)at the Endoscopic Treatment Center of the First Affiliated Hospital of China Medical University.This study was approved by the Ethics Committee of the First Affiliated Hospital of China Medical University,and all subjects have signed informed consent.2.ST-seq library preparation,staining,and imaging: Four spatial libraries were constructed using a ST-seq library preparation microarray.FFPE tissues were cut by microtome with a thickness of 10 μm and installed on gene expression slides for HE staining and subsequent imaging scanning.3.Permeation,reverse transcription,and c DNA amplification: Spatial gene expression was processed using Visum spatial gene expression slides and reagent kits.c DNA amplification was performed on a S1000 TM touch thermal cycling apparatus(Bio Rad).4.Construction and sequencing data reading of spatial gene expression library:Construction of spatial gene library using Visum spatial library construction kit.After qualified library,Illumina Novaseq6000 sequencer was used for 3 ’end sequencing.5.Spatial transcriptome data processing: We used Space Ranger software to deal with the FASTQ raw files and histological image,with default parameters.Filtered gene point matrix and baseline aligned low resolution images were used for Seurat analysis of downlink data.6.Hierarchical clustering was performed for each sample: Seurat package was used for gene expression normalization,dimension reduction,spot clustering and differential expression analysis.7.Comprehensive analysis of multiple samples: The Seurat package was used for comprehensive analysis of 4 samples.8.Functional enrichment analysis and protein-protein interaction analysis: Enrichment tests of candidate gene sets based on hypergeometric distribution were calculated using the cluster Profiler R package,with GO and KEGG pathways as reference databases.Protein-protein interaction(PPI)was derived from String database data using the binding_score≥400.The interactions of the top 50 genes in each cluster were extracted from the database.PPI results were visualized using Cytoscape software.9.st Learn spatial trajectory analysis: First the PAGA algorithm was used to probe the connections between groups.The diffusion pseudotime method was used to calculate the pseudo-time.Then,the pseudo-space-time distance(PSTD)was calculated by combining the gene expression and physical distance.10.Statistical analysis: R(http://www.r-project.org)and SPSS 19.0 were used for all statistical analysis.Part Ⅱ.Correlation between OLFM4 expression and risk of GC and clinicopathological parameters1.Subjects: This study was approved by the Ethics Committee of the First Affiliated Hospital of China Medical University,and all subjects have signed informed consent.The 107 GC samples in this study were hospitalized in the Department of gastroenterology(27 cases)and the anorectal surgery department(80 cases)between December 2014 and October 2020 in the First Affiliated Hospital of China Medical University,and underwent endoscopic submucosal dissection(ESD)and subtotal gastrectomy in patients with GC.Inclusion criteria: no preoperative chemoradiotherapy,immunotherapy,etc.;without other primary tumors;GC and para-cancer tissue samples obtained through endoscopic or surgical procedures,and confirmed as GC by independent pathological diagnosis by two senior pathologists.The epidemiological data of inpatients,including drinking history and smoking history,were obtained by inquiring inpatient medical records.Clinicopathologic information was obtained by querying the pathological diagnosis report,and TNM clinical staging of patients was performed according to the TNM staging of GC in the seventh edition of UICC/AJCC.2.Immunohistochemical staining and immunohistochemical scoring were performed:the expression of OLFM4 in different tissues was scored separately by two pathologists.The final immunostaining score was the product of staining intensity and staining area score.3.Bioinformatics analysis: The TCGA-STAD(GC)data collected in this study were downloaded from UCSC Xena,and rank sum test was used to determine the differential expression of OLFM4 m RNA in gastric cancer tissues(P<0.05).Analysis and visualization were performed by RStudio 3.6.1.UALCAN database was used to analyze the relationship between OLFM4 and different clinicopathological parameters.Prognostic analysis was performed using Kaplan-Meier Plotter and UALCAN databases.4.Statistical analysis: Chi-square was used to test the relationship between OLFM4 protein expression and clinicopathological parameters.Wilcoxon rank sum test was used to analyze the difference between 2 groups.Statistical tests were all analyzed using SPSS software(v19.0),and bilateral P<0.05 was considered to be statistically significant.The Graph Pad software(v8.0)was used for data processing and mapping.Part III: Effects of OLFM4 expression on biological behavior of GC cells and its mechanisms1.The in vitro study used AGS GC cell line and HGC27 GC cell line.Cells were purchased from Cell Resource Center,Institute of Basic Medicine,Chinese Academy of Medical Sciences,with STR identification certificate.2.OLFM4 stable overexpressed cell lines were constructed,and fluorescence was observed 72 hours later for resistance screening3.Total RNA of cells was extracted by phenol-chloroform method and converted into c DNA.m RNA expression levels in cells were detected by RT-q PCR.4.The protein was extracted by adding phosphatase and protease inhibitor with NP40.Protein quantification was performed using an enzyme-labeled instrument,and the protein samples were then used for western blotting.5.CCK8 reagent and EDU fluorescence staining were used to detect cell proliferation activity.6.Transwell migration assay was used to detect the migration ability of HGC27 cells and AGS cells.Transwell invasion assay was used to detect the invasion ability of HGC27 cells and AGS cells.7.SPSS software,version 19,was used for statistical analysis of data used in this part of the study.All experiments were independently repeated for more than three times,and the values of measurement data were expressed as mean ± standard deviation or standard error.Paired sample t test and nonparametric test were used for statistical analysis of data results.P value less than 0.05 was considered statistically significant.The Graph Pad software(v8.0)was used for data processing and mapping.Results:Part Ⅰ.Construction of spatial transcriptome maps at different stages of EGC progression1.Tissue array preparation and basic sequencing information of samples.In this study,each of the 8 FFPE samples was punched for two areas: one covering both superficial gastritis(GS)and IM tissues,and the other covering both IM and GC tissues.The corresponding FFPE tissues were punched for the selected area,and two samples with the same location and histological classification were re-embedded into a paraffin arrays.The eight samples were re-embedded into four paraffin arrays,and then were performed by 10 x Genomics platform for ST-seq.Transcriptome data of6913 loci were obtained from 4 sequencing arrays,and the sequencing depth of a single spot was 1987-4016 genes.2.Cluster visualization and pathologic identification of gene expressionPrincipal component analysis(PCA)was used to conduct dimensionality reduction clustering for high-variation genes of these spots,and 8-9 different groups/tissue regions were generated.The clustering results were visualized by UMAP and t-SNE.After pathologist identification,the groups were defined as GS/Normal(N)group,IM group,GC group,muscularis mucosae(MM)group and submucosa(SM)group,etc.The first 10 specific expression genes of each group were shown by heat map.3.Construction of spatial transcriptome maps at different stages of GC progression3.1 Characteristics of GS spatial transcriptome map3.1.1 GS clusters and GS specific genes in different clustersAccording to gene expression and spatial location characteristics,GS tissues of antrum were subdivided into 2 groups,which were defined as pit epithelium group and glandular group respectively.The GS tissue of the body of the stomach was divided into 3 groups,which were defined as the pit epithelium group,the neck of the gland group and the gland group.These group-specific genes were calculated by difference analysis,as shown in volcano map and heat map.GKN2,TFF1 and CA2 were highly expressed in the pit epithelium of antrum,while MUC6,LIPF and PGC were highly expressed in the glands of antrum.GKN1,GKN2,and PHGR1 were highly expressed in the pit epithelium of the stomach body,HIST1H2 AD,LYZ,and KCNE3 were highly expressed in the glandular neck,and SST,GAST,and LTF were highly expressed in the gland.3.1.2 Functional enrichment analysis of GS specific genes in different subgroupsIn order to explore the biological function of each group of specific genes,we performed GO,KEGG and Reactome functional enrichment analysis.The results showed that the antral pit epithelium was mainly involved in adhesion,retinol metabolism,cytochrome P450 metabolism of exogenous drugs and other pathways.The antral glands are mainly involved in extracellular matrix receptor interactions,the plaque kinase pathway,and the TGF-β pathway.The pit epithelium of the stomach is mainly involved in extracellular matrix receptor interaction and plaque kinase pathway.Glandular neck group is mainly involved in cell cycle,DNA replication and proliferation-related pathways.Glandular genes are involved in metabolism,oxidative phosphorylation,gastric acid secretion and other pathways.3.1.3 Interaction network of GS specific gene proteins in different subgroupsThe PPI network was further constructed and the key genes were screened.The results showed that the key genes were NQO1,AKR1C1,GPX2,etc.The key genes of antral glands were FN1,CLU,CD44,etc.The key genes were LAMA3,LAMB3,ITGA5,etc.The key genes were HIST1H4 A,HIST1H4D,HIST1H4 E,etc.The key genes of gastric gland were CYC1,COX5 B,NDUFB9 and so on.3.2 Analysis of spatial transcriptome characteristics of IM3.2.1 IM clusters and IM specific gene analysis in different clustersThe IM tissues obtained in this study were para-carcinomatous of EGC.MME/CD10(small intestine brush border marker)is known to be specifically highly expressed in CIM,while MUC5 AC or MUC6(gastric epithelial marker)is specifically highly expressed in IIM.As a marker of goblet cell,the main component of IM,MUC2 was expressed in both types of IM.According to above principles,seven IM groups could be defined as one CIM(IGCb-IM2)and six IIMs.Differential gene analysis was performed for CIM and IIMs and the specific expressed genes were shown on volcano plot,among which APOA1,APOC3,SLC46A3,APOA4 and ABCC2 were specifically expressed in CIM,and GKN1,GKN2,LIPF,PGC,TFF2,KRT7 were specifically expressed in IIM.3.2.2 Functional Enrichment Analysis of IM characteristic Genes in different groupsIn order to explore the biological functions of specific genes in each clusters,GO,KEGG and Reactome functional enrichment analysis were performed.The results showed that CIM specific genes were mainly involved in the formation of brush border,lipid metabolism,digestion and absorption of fat,protein and vitamin,and PPAR signaling pathway.The enrichment analysis of the other 6 cases of IIM confirmed that the occurrence of IIM may be related to various pathogenic mechanisms.The IIM1 group(IGCa-IIM1)specific genes were mainly involved in focal adhesion,PI3K-Akt signaling pathway and proteoglycan in cancer.IGCa-IIM2 group specific genes were mainly involved in mineral absorption and cell response to metal ions such as copper and zinc.IGCb-IIM1 genes are mainly involved in the activation of PI3K-Akt and IL-17 signaling pathways.The expression of specific genes of DGCa-IIM1 group might be related to metabolism,hepatitis,EBV virus,influenza A virus infection,etc.The role of DGCa-IIM2 were in the invasion of epithelial cells by pathogenic Escherichia coli,Salmonella and other bacteria;DGCb-IIM specific genes mainly participated in alcoholism and viral carcinogenesis.3.2.3 Protein-protein interaction network analysisProtein-protein interaction(PPI)network analysis of specific genes in different clusters of IM was further constructed,which showed that seven hub genes APOB,APOA4,APOA1,APOC3,APOL1,FABP1,MTTP appeared in CIM cluster and three hub genes ISG15,HLA-A,and OASL in IIM cluster.3.2.4 Spatial evolution trajectory during IM differentiationIn order to explore the development process of gastric diseases,we applied PSTD method to analyze the possible evolutionary trajectory between clusters.Spatial trajectory analysis showed that the "glandular epithelium" region of GS might preferentially evolve into the proliferative region of IM which the stem-like cells(OLFM4+EPHB2+ cells)of IM and the proliferative region of IM located.Then,cells in the IM proliferating zone moved upward and differentiated into intestinal epithelial cells(FABP1+ cells)or goblet cells(MUC2+ cells).3.3 Characteristics analysis of EGC spatial transcriptome map3.3.1 EGC clusters and EGC specific genes in different clustersIn order to analyze the ST heterogeneity of EGC,four EGC clusters(IGCa,IGCb,DGCa,DGCb)with different histological classification and location were collected.IGCa and IGCb were diagnosed as high-differentiated EGCs located in the antrum and body of stomach,respectively.DGCa was poor-differentiated EGC in the antrum and DGCb was signet-ring cell carcinoma of the stomach body.Top 10 specific genes were exhibited on heat map.3.3.2 Functional Enrichment analysis of EGC specific genes in different clustersFunctional analysis showed that DGCa cluster-specific expression genes were associated with viral infection and innate immunity,cell polarity,migration and cell cycle progression,regulation of differentiation,and DNA methylation.DGCb cluster-specific expression genes were enriched in cancer pathways that influence cell polarity,proliferation and migration,immune-inflammatory processes,and cell adhesion regulation.Specifically expressed genes of IGCa cluster were involved in cancer pathways,such as PI3K-Akt signaling pathway,gastric cancer pathway,and cell response to metal ions/mineral absorption pathway.IGCb gene was related to lipid metabolism,post-translational protein modification,TCA cycle and respiratory electron transport,etc.3.3.3 The PPI network construction for EGC specific genes for different clustersThe hub genes were found through cytoscape software.The results showed that the hub genes of IGCa cluster were EEF1B2、EEF2、GNB2L1、EEF1G、PPP1R1B、MT1E、MT1G、MT1X、MT1H、LGR5.The hub genes of IGCb group were MUC4、MUC3A、MUC5B、MUC13、IL1B、NOS2、CXCL3、ASS1、DGAT1、UQCRQ.The hub genes of the DGCa group were HIST1H4B、HIST1H3I、HIST1H4D、HIST1H4A、HIST1H3D、HIST1H3G、HIST1H2BI、HIST1H2AD、HIST1H1E、HLA-A.The hub genes of DGCb group were DNAJB1、HSPH1、DNAJA4、HSPA8、HSPA6、HSPA1A、HSPD1、BAG3、HSP90AA1、HSP90AB1.4.Different expression patterns of genes during EGC progressionGenes were divided into four different expression patterns(DEPs)according to their expression at different stages.The genes of type Ⅰ expression model(M1)were mainly expressed in GS stage,but decreased in IM stage and EGC stage(GS>IM≥EGC).The type Ⅱ expression model(M2)genes were mainly expressed in IM expression level(GS<IM>EGC).The type Ⅲ expression model(M3)was significantly reduced in GS,and overexpressed in IM and EGC.The expression model of type Ⅳ(M4)showed high specific expression in EGC.Additionally,using M1-M4 gene set,we performed ss GSEA enrichment scores for IM and EGC clusters.The results showed that M2 was mainly enriched in CIM group,while M1 and M4 were enriched in GS group and EGC group,respectively.M3 was enriched in both IM and EGC,indicating that it might pre-warn the occurrence of EGC.Therefore,our subsequent study focused on the function and effect of OLFM4 which was the top1 gene in M3.Part Ⅱ: Correlation between OLFM4 expression and the GC risk and clinicopathological parameters1.Correlation between OLFM4 expression and the risk of GCIn this study,107 gastric cancer tissues were collected,including 107 pairs of cancers and their corresponding para-cancer normal/IM tissues.Immunohistochemical results illustrated that OLFM4 was rarely expressed in normal tissues,but highly expressed in IM and GC tissues.We also analyzed OLFM4 m RNA expression in GC and para-cancer tissues using the TCGA database,and the results showed that OLFM4 expression in GC was significantly higher than that in para-cancer tissues.These results indicated that OLFM4 expression was associated with the risk of gastric cancer.2.Correlation between OLFM4 expression and clinicopathological parameters of GCAccording to the expression of OLFM4 in GC,the high expression and low expression groups were defined.Further analysis showed that the expression of OLFM4 was associated with the gross type(P<0.001),histological classification(P=0.001),depth of invasion(p T grade)(P=0.006),vascular cancer thrombus(P=0.033)of GC.The high expression group of OLFM4 has higher proportion of EGC,higher proportion of IGC and higher proportion of T1+T2 stage than low expression group.Furthermore,UALCAN database(http://ualcan.path.uab.edu/index.html)was applied to analyze the association of OLFM4 m RNA expression with clinical pathological parameters of GC.The results showed that OLFM4 expression was higher expressed in male GC patients and in well-differentiated GC.There was no statistical difference between helicobacter pylori-infected or not and TP53 mutated or not.3.Correlation between OLFM4 expression and prognosis of gastric cancerIn order to investigate whether OLFM4 affects the prognosis of patients with gastric cancer,we used Kaplan-Meier Plotter and UALCAN database to analyze the relationship between OLFM4 expression and overall survival rate of gastric cancer.Patients with high OLFM4 expression had a higher cumulative survival rate than those with low OLFM4 expression(95%CI:0.69-0.97,HR=0.82,P=0.024).Part III: The effect of OLFM4 expression on biological behavior of GC cells and its mechanism1.Effects of OLFM4 expression on biological behavior of GC cells1.1 Construction of lentivirus stable cell lineWe transfected OLFM4 overexpressed lentivirus into HGC27 and AGS gastric cancer cell lines.The transfection efficiency was observed 48 h later by fluorescence microscopy and verified by RT-q PCR.The results showed that the efficiency of OLFM4 over-expression lentivirus in HGC27 and AGS gastric cancer cell lines was more than 200 times.1.2 Effects of OLFM4 expression on proliferation of GC cellsWe tested the effect of OLFM4 on the proliferation of GC cells using CCK-8 test.CCK-8 reagent was added to OLFM4 over-expressed GC cell lines and NC cell lines to detect the cell activity at 0 h,24 h,48 h and 72 h,and the cell proliferation curves of the two groups were drawn by Graphpad.The results of CCK-8 showed that OLFM4 could inhibit the proliferation of gastric cancer cells HGC27 and AGS.The Edu proliferation detection kit was further used to detect the effect of OLFM4 on the proliferation of gastric cancer cells HGC27 and AGS.It was found that the proportion of positive Edu was reduced after over-expression of OLFM4,which again indicated that OLFM4 could inhibit the proliferation of gastric cancer cells.1.3 Effects of OLFM4 expression on migration and invasion of GC cellsWe tested the effect of OLFM4 on the migration of gastric cancer cells HGC27 and AGS by transwell migration assay.The results showed that over-expression of OLFM4 significantly reduced the migration ability of gastric cancer cells HGC27 and AGS.Further,transwell invasion assay was used to detect the effect of OLFM4 on the invasion of gastric cancer cells HGC27 and AGS.The results showed that the invasion ability of gastric cancer cells HGC27 and AGS was also weakened after over-expression of OLFM4.1.4 Regulation of OLFM4 expression on cuproptosis in gastric cancer cellsIn order to further explore other important functions that OLFM4 may be involved in besides the classical malignant biological functions of tumors,We used ST-seq to perform GO and KEGG functional enrichment analysis of OLFM4 co-expressed gene sets.In addition,we also used the data from the TCGA-STAD database to analyze the GSEA functional enrichment analysis of genes positively correlated with OLFM4.ST data and TCGA data analysis results jointly indicate that OLFM4 mainly plays a role in the regulation of intracellular copper ion homeostasis.In this study,the regulation of OLFM4 on intra-cellular copper ions was studied by constructing intra-cellular copper accumulation in vitro.Firstly,the cell viability was explored according to the copper ion concentration recommended in the literature of other tumor cells.We added 160μmol,320μmol and 480μmol Cu SO4 into the cell medium respectively,and tested the cell viability with CCK-8.The results showed that the cell viability was reduced by nearly 40-50% under the exposure of 320μmol/L Cu SO4.Subsequently,according to the results of the latest literature on cuproptosis,WB experiment was performed on HGC27-OLFM4 overexpressed cells stimulated by 320μmol Cu SO4 concentration and its NC to observe whether there were differences in the expression of cuproptosis protein between the two groups.The results showed that the expressions of LIAS,FDX1 and DLAT,which are key factors in cuproptosis were decreased.2.Mechanism of OLFM4 expression on biological behavior of GC cells2.1 OLFM4 inhibits ITGA2/FAK/AKT pathways in GC cellsWe has illustrated that OLFM4 could inhibit the proliferation,invasion and migration of GC cells.However,we needed to further explore the mechanisms of the influence on biological behavior of GC cells.Studies had shown that FAK could receive upstream integrin signals and activate downstream intracellular PI3K/Akt signaling pathway to promote cell proliferation and invasion.In this study,the upstream and downstream of FAK were explored,and OLFM4 was found to inhibit the expression of integrin-ITGA2 and the activation of p-Akt.2.2 OLFM4 affects the cuproptosis of GC cells by regulating the expression of ATP7 B.In order to explore the cause of OLFM4’s influence on cuproptosis,this study analyzed the influence of OLFM4 on ATP7 B,which maintains copper ion homeostasis in cells.The results showed that OLFM4 could reduce the expression of ATP7 B at m RNA level and protein level under the condition of normal 1640 complete culture medium.In addition,WB experiment was also used to analyze whether the expression regulation relationship still existed after exposure to 320μmol/LCu SO4 for 6 hours,and the results showed that OLFM4 could still reduce the expression of ATP7 B.Conclusions:1.In the present study,we constructed the spatial transcriptome Atlas of different stages of GC progression,revealed the specific genes,enrichment pathways and protein interaction networks of IM and EGC in different groups based on space site,as well as the spatial evolution tracks of IM differentiation,and established four gene sets of different expression patterns during development of GC.2.OLFM4 was a hub gene of the pre-warning GC gene sets.The expression of OLFM4 in IM and GC were significantly higher than GS.The high expression of OLFM4 was negatively correlated with the gross types of GC,invasion depth and vessel carcinoma embolus,indicating a good prognosis for patients with GC.3.OLFM4 had inhibitory effects on proliferation,migration and invasion of GC cells,and could promote cuproptosis under certain conditions.4.OLFM4 might affect the malignant biological behavior of gastric cancer cells through ITGA2/FAK/AKT pathway.5.OLFM4 might promote cuproptosis by downregulating ATP7 B protein. |
PDF Full Text Request