| Gastric adenocarcinoma is the most common type of gastric cancer,and more than 90% of gastric cancer cases are gastric adenocarcinoma.It is known that the pathological characteristics of gastric cancer are affected by a variety of genetic and environmental factors.As a result,gastric adenocarcinoma exhibits a high degree of morphological and molecular heterogeneity,and this heterogeneity hinder the accurate diagnosis of gastric cancer and poses a series of challenge in clinical treatment,throughout management,and prognostic assessment.With the development of high-throughput sequencing technology and in-depth analysis of tumor biological phenomena,it has become particularly important to analyze the heterogeneity of gastric cancer from the molecular level.Conventional tumor sequencing(bulk)sequencing can detect the average expression level of genes,masking the characteristics of different cell populations,so it is impossible to analyze cell diversity and heterogeneity of cell transcriptome status in tissues.In addition,studies that label and purify specific cell populations with known antibodies cannot cover all cell populations,and cannot detect rare cell populations or cell populations with specific phenotypic states.Therefore,for the analysis of the heterogeneity of gastric cancer,there is an urgent need for a comprehensive and systematic analysis of the expression of single cells in gastric pre-neoplastic and malignant mucosa.The rapid development and joint application of single-cell sequencing technology and microfluidic chip technology allow researchers to systematically analyze the cellular components in tissues and the transcriptome characteristics of individual cells at a single-cell resolution without bias.This study intends to use this high-throughput,high-resolution transcriptome sequencing technology to conduct a comprehensive and systematic cellular and molecular census of normal gastric tissues,chronic gastritis gastric mucosa tissues and gastric cancer mucosa tissues to explore the origin of metaplastic cells and analysis of gastric cancer heterogeneity.This study included 3 cases of non-malignant gastric mucosa tissue and 9 cases of primary gastric adenocarcinoma tissue.Fresh samples were divided into two equal portions,one part is dissociated and digested into a single-cell suspension for sorting live cells for single-cell sequencing,and the other part is used for other experiments,including EBER detection and immunofluorescence.After strict data quality control,more than 26,766 single cell was sequenced and the transcriptome profiles were obtained.The transcriptome expression matrix of these 26,766 single cells was normalized,and the principal component dimensionality reduction analysis was performed.The first 10 principal components were selected for t SNE dimensionality reduction and visualization.At the same time,by analyzing the expression matrix of TCGA dataset including gastric cancer tissues and normal tissues samples,the gastric cancer related characteristic genes were screened and iterative discriminant analysis of benign and malignant gastric epithelial cells was performed based on scoring of these genes.After identifying the benign and malignant gastric epithelium,we analyzed the cell composition and molecular characteristics of non-malignant gastric epithelium.We found six types of gastric epithelium: gastric chief cells,parietal cells,endocrine cells,metaplastic cells,neck cells,and surface pit cells.By analyzing the gene expression characteristics of gastric chief cells,metaplastic cells and neck cells,we found that gastric chief cells derived from gastritis mucosa are closely related to gastric metaplastic cells by unbiased developmental trajectory analysis.The pseudotime trajectory analysis shows that MUC6 and TFF2 expressing gastric metaplastic(spasmolytic polypeptide-expressing metaplasia,SPEM)cells are derived from the gastric chief cells and confirms the concept that the gastritis metaplastic cells can originate from the trans-differentiation of the gastric chief cells.Based on single-cell transcriptome findings,we provide molecular evidence of the origin of gastric metaplastic cells,which will facilitate the understanding of gastric precancerous lesions.Gastric cancer is known to be a highly heterogeneous disease with different histopathological and molecular characteristics.We re-dimensioned the gastric malignant epithelial cells and performed cluster analysis to obtain 5 different groups(C1-C5).We found that these groups were gastric cancer histological types or patient-specific,indicating a high degree of inter-tumor heterogeneity.Among them,the diffuse type sample(C1)highly expresses gastric characteristic gene CLDN18,but differentiation-related genes KRT20,PHGR1,MUC13 have low expression level.While the intestinal type sample(C2)exhibited a low expression level of CLDN18 transcript,but these differentiation related gene expression levels vary among patients and cells.The cells from mixed-histology sample(C3)did not have very specific characteristic genes,but co-expressed genes related to diffuse and intestinal cells.Interestingly,we found two different types of differentiation modules in the two intestinal samples,suggesting that gastric malignant cells have high intra-tumor heterogeneity in the same individual.We scored the differentiation degree of these malignant cells,and we found that the diffuse-type samples showed a low degree of differentiation,while the intestinal-type samples exhibited various differentiation degree.Furthermore,according to the selected differentiation-related genes,we scored the differentiation of gastric cancer samples in TCGA dataset and found that the tumor differentiation degree is highly correlated with the prognosis of gastric cancer.We found two groups that are different from the traditional Lauren type,in which cells in the C4 group highly express gastric main cell markers(such as LIPF and PGC)and are related to the Wnt/ β-catenin signaling pathway.We call this subgroup-dominated tumor the predominantly cell-differentiated adenocarcinoma(GA-FG-CCP)and confirm the presence of GA-FG-CCP using transcriptomic features and histological images in two public batch data sets.Another subgroup is infected with Epstein-Barr virus(EBV),which specifically expresses immune-related signature genes(such as LY6 K and MHC-II)and enrich antigen processing and presenting pathway.Considering that MHC-II molecule is a relevant target for tumor immunotherapy,our single-cell transcriptome data may reveal the reasons why patients with EBV-positive gastric cancer benefit from immunotherapy.In summary,we constructed a systematic and unbiased single-cell transcriptome profile of gastric cancer,which included 26,766 cells from 9 gastric cancer samples and 3 adjacent non-malignant gastric mucosa sample.Focusing on epithelial cells,we identified a panel of biomarkers that distinguish benign and malignant epithelium.The results of potential gastric chief cells transition to metaplastic cells should help elucidate the mechanism of gastric cancer genesis.We also characterize differentiation degrees within and between tumors and predict poor prognosis in patients with poorly differentiated degree.Furthermore,we identified GA-FG-CCP and EBV+ gastric cancer using both sc RNA-seq and bulk TCGA dataset.Because this is a preliminary study,the limitation of our analysis is the small number of patient cases participating in the study,and our findings need to be validated in a larger patient cohort.However,despite this,GA molecular profiling,characterized by single-cell transcriptome profiling,may pave the way for dissecting tumor heterogeneity and is of great significance for clinical practice. |
PDF Full Text Request