Single-cell transcriptomic profiling uncovers cellular complexity and microenvironment in gastric tumorigenesis associated with Helicobacter pylori

doi:10.21203/rs.3.rs-3641851/v1

Download PDF

Research Article

Single-cell transcriptomic profiling uncovers cellular complexity and microenvironment in gastric tumorigenesis associated with Helicobacter pylori

https://doi.org/10.21203/rs.3.rs-3641851/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background Helicobacter pylori (H. pylori) infection is a key initiating factor in the Correa cascade of gastric carcinogenesis, but the comprehensive understanding of the pathogenic mechanisms underlying H. pylori -induced gastric cancer (GC) remains elusive. Here, we generated a single-cell atlas of gastric tumorigenesis comprising 18 specimens of gastritis, intestinal metaplasia and GC with or without H. pylori infection.

Results Gradually decreased proportion of epithelial cells and increased percentages of myeloid cells were observed in Correa carcinogenic cascade. Remarkably, the epithelial subset enterocytes were highly enriched in intestinal metaplasia, with high expression of lipid metabolism-related genes, and HNF4G was predicted as its specific transcription factor. Compared to non-cancerous tissues, GC exhibited completely different fibroblast cell states, with enrichment of immune regulatory functions. Furthermore, the differentially expressed genes, in H. pylori-positive and negative epithelial cells, fibroblasts and myeloid cells, were identified. Notably,the stromal cells, particularly inflammatory-cancer associated fibroblasts (iCAFs) and tumor-like Endothelial cells (ECs), play an important contributory role in GC microenvironment. H. pylori-positive specimens exhibited enriched cell-cell communication, with significantly active SPP1, TNF signaling networks.

Conclusions Our study provides a comprehensive landscape of gastric carcinogenesis ecosystem and novel insights into the molecular mechanisms underlying H. pylori-induced GC.

ScRNA-seq

cellular heterogeneity

microenvironment

H. pylori

gastric carcinogenesis

Gastric cancer (GC) is the fifth most common malignant tumor and the third leading cause of cancer-related deaths worldwide, particularly in East Asian countries(1). The majority of GC cases are diagnosed at an advance stage(2). Helicobacter pylori (H. pylori), a gram-negative bacterium that persistently colonize the gastric mucosa, is the widely-recognized known risk factor for GC. In fact, the World Health Organization (WHO) has classified H. pylori as a Group I carcinogen leading to gastric adenocarcinoma(3). The classical Correa cascade of multistep gastric carcinogenesis generally involves a series of precancerous lesions, where chronic gastritis develops through metaplasia, dysplasia to ultimately intestinal type GC. H. pylori infection plays an initiating causative role in this process(4, 5). Although the benefit of H. pylori eradication in reducing GC risk has been reported by extensively clinical evidence(6), the issues become complicated if precancerous lesions have appeared. A thorough exploration of gastric tumorigenesis associated with H. pylori would help in understanding of the molecular mechanisms and discover more effective strategies for GC diagnosis and treatment.

Multiple omics studies on genomics, transcriptomics and proteomics elaborated various molecular landscapes underlying the progression of GC(7, 8). However, GC is a complex ecosystem, defined by the heterogeneity and interactions between diverse cell populations, including epithelial cells, stromal cells and immune cells. Single-cell RNA sequencing (scRNA-seq), a powerful tool to assess gene expression at the individual cell level, has been recently used to explore gastric carcinoma microenvironment with particular focus on the transcriptional patterns in epithelial and fibroblast cells(9–11). However, cellular diversity and cell-cell communications in the context of gastric tissue microenvironments during the Correa cascade of multistep gastric carcinogenesis remains poorly understand. Given the causal determinant role of H. pylori in gastric inflammation and disease progression, extensive evidence focuses on the molecular biological and immune regulation of gastric epithelial cells and myeloid cells. H. pylori infection has been indicated to endow epithelial cells with multiple phenotypes, such as sustained inflammation, hyperproliferation, and invasiveness(12, 13). Additionally, adhesion of H. pylori triggers double-strand breaks (DSBs) and elicits genome instability in epithelial cells(14). The transcriptional profiles of other cell types such as fibroblasts and intercellular communication are still lacking.

In this work, we dissected cellular diversity and microenvironment heterogeneity by analyzing 91394 single-cell transcriptomes from 18 samples of non-atrophic gastritis (GS), intestinal metaplasia (IM), and GC that were either H. pylori positive or negative. We observed reduction of epithelial cells and increased fractions of endothelial cells (EC) and myeloid cells in GC tissues, compared with those in GS or IM tissues. Furthermore, we identified and functionally validated a unique subpopulation of epithelial cells, known as tumor-like cells, which were enriched in GC tissues, as well as a distinct epithelial subset known as enterocyte cells that were derived from the precancerous IM tissues. Additionally, we explored the heterogeneity and molecular process of stromal cells, including fibroblasts and ECs, with further determination of the interactions between these cell populations in gastric carcinogenesis. We also uncovered the transcriptome features of tumor-associated macrophage (TAM), a subtype of myeloid cells that were abundant in GC tissues. Notably, the differentially expressed genes (DEGs) in epithelial cells, fibroblasts and myeloid cells were identified from H. pylori-positive and negative samples, which were further validated by using other independent cohorts and transcriptomic data from GEO database. Our data sheds light on the molecular characteristics and potential intercellular signals controlling cell lineages states, which provides a comprehensive understanding of the pathogenic mechanisms of H. pylori and gives the path to development of new strategies for precise prevention and therapeutics of gastric carcinoma.

ScRNA-seq landscape along the consecutive gastric carcinogenesis associated with H. pylori

To explore the cellular heterogeneity and microenvironment in gastric tumorigenesis associated with H. pylori infection, we collected 18 gastric specimens representing the progressive histologic stages, i.e., 6 cases of GS, 6 cases of IM, and 6 cases of GC specimens (Figure 1A). Each stage of the lesion was divided into two groups based on H. pylori infection status. Detailed patient information, including age, sex, pathological information, H. pylori infection status, was provided in Table S1. Tissue samples were dissociated into single-cell suspensions and processed using the 10× Genomic Chromium workflow (Figure 1B).

After data quality control and filtering, a total of 91394 cells were used for further analysis (2563-7677 cells; median 5077 cells/patient) (Table S1). The resulting 26 cell clusters were annotated as 9 cell types by established marker genes(10, 15) (Figure 1C and Figure S1A). Non-immune cells included epithelial cells (PECAM, KRT19), ECs (PECAM1, CDH5), fibroblasts (COL1A1, PDGFRA), and smooth muscle cells (TAGLN, ACTA2); immune cells included T cells (CD3D, CD3G), B cells (CD79A), myeloid cells (CD68, CSF1R), plasma cells (MZB1, IGKC), and mast cells (TPSAB1, TPSB2) (Figure 1C; Figure S1B and Table S2). Epithelial cells expressed high levels of gastric mucosa-specific genes MUC5AC and TFF1 (16)(Figure S1C), as expected, indicating the accuracy of our data.

Gradually decreased proportion of epithelial cells, accompanied by increased percentages of myeloid cells during gastric carcinogenic cascade.

All of these cell types were present among patients, albeit in varying proportions (Figure S1D). When comparing the cellular compositions across the three progressively pathological stages (GS, IM and GC), we found that the GS samples were dominated by epithelial cells and T cells. Notably, the epithelial cells gradually decreased in proportion, accompanied by gradually increased myeloid cells as the disease progressed to cancer. Additionally, the ECs were greatly expanded in cellular proportion and significantly increased in GC tissues, compared to GS and IM tissues (Figure 1D). Furthermore, H. pylori-infected samples contained relatively lower proportion of epithelial cells compared to uninfected tissues, while the trend was opposite for ECs and myeloid cells (Figure 1E), in line with the RNA expression levels of the marker genes of these three cell types in bulk RNA-seq data in gastric tissues (Figure S1E). The top 10 candidate marker genes for each cell types were shown in Figure 1F. We further performed the IHC staining to validate the changes of endothelial, myeloid and epithelial cells during the cascade of gastric carcinogenesis. Consistently, upregulation of marker genes CD31 and CD68 for ECs and myeloid cells, respectively (Figure 1G), and downregulation of gastric epithelial cell-specific marker TFF1 were observed in GC tissues, compared to GS and IM specimens (Figure 1H). In addition, the increased expression levels of CD68, as well as the decreased levels of TFF1, were also supported by TCGA gastric carcinoma cohort (Figure S1F). The analysis of several public datasets from GEO database (GSE122401, GSE65801 and GSE13195) by comparing cancer tissues and paired normal adjacent tissues also showed consistent data (Figure S1G and H). These differences highlight the dynamic changes of cellular compositions in gastric tumorigenesis.

Distinct cell lineage and transcriptional patterns of epithelial cells during multistep gastric carcinogenesis

To explore lineage states and transcriptional changes in gastric tumorigenesis, we next performed unsupervised dimensionality reduction of epithelial cells. A total of 19 clusters emerged, including pit mucous cells, neck mucous cells, chief cells, parietal cells, endocrine cells, enteroendocrine cells, enterocytes, and tumor-like epithelial cells (tumor-like EPC) (Figure 2A and Figure S2A). Several common epithelial clusters were identified using known marker genes, such as MUC5AC and TFF1 (cluster 1,3 and 4) representing pit mucous cells, MUC6 representing neck mucous cells expressing (cluster 5 and 13), PGA3 and PGA4 representing chief cells (cluster 6), and ATP4A and ATP4B representing parietal cells (cluster 8 and 18)(17, 18). Endocrine cells (cluster 2, 9, 16, and 17) and enteroendocrine cells (cluster 11, 12, 14 and 19) were identified based on high expression of GAST and SST, CHGA and TPH1, respectively(19) (Figure 2B; Figure S2B). They exhibited clearly different states from other cell lineages. As expected, the classical epithelial cell populations, such as pit mucous cell, neck mucous cells, parietal cells, showed significantly decreased cell numbers across gastric carcinogenesis cascade (Figure S2C). Interestingly, epithelial cluster 10 expressing FABP1 and PHGR1, defined as enterocytes, were almost derived from the precancerous tissue of gastric intestinal metaplasia. Additionally, cluster 7 named tumor-like PEC, expressing cancer embryonic antigen CEACAM5, were enriched in GC tissues ((Figure 2C and Figure S2D).

Given that cancer cells often occur multiple mutations and chromosomal alterations, we further utilized copy number variants (CNV) analysis to classify epithelial cells as either non-malignant or malignant cells. As expected, tumor-like EPC displayed higher CNV levels compared to other cell types (Figure S2E). We also compared the CNV scores of epithelial cells across all samples, and found it to be significantly elevated in GC tissues, particularly in H. pylori -positive subgroups, compared with GS or IM tissues (Figure 2D). Due to the high proportion of tumor-like EPC cluster in GC samples, we performed multicolor IF staining and verified its significant abundance (EPCAM⁺/CEACAM5⁺) in gastric carcinoma tissues (Figure 2E). We noticed that KRT17 was extremely enriched in tumor-like EPC and in GC tissues (Figure S2B and F), suggesting its potential cancer-promoting role. IF and IHC staining consistently confirmed the epithelial cell-specific expression of KRT17 and its significant increase in GC tissues (Figure 2F), which were further confirmed by GC-related GEO datasets and TCGA cohort with primary tumor specimens and matched normal specimens (Figure S2G). We next aimed to explore the cellular origin and transitions of tumor-like EPC population. We surprisingly found that this cell subtype simultaneously expressed MUC5AC and TFF1, the classical pit mucous cell markers. We validated the presentence of MUC5AC in tumor-like EPC (MUC5AC⁺CEACAM5⁺) from GC samples by multicolor IF staining (Figure 2G). In line with these data, trajectory plot analysis showed that pit mucous cells were the origin of tumor-like EPC (Figure 2H). To characterize the functions and activation pathways of tumor-like EPC, we performed a gene set enrichment analysis (GSEA). Intriguingly, the gene set of tumor-like EPC was enriched in well-known pro-cancer signaling pathways, such as Notch, PI3K/AKT, and Wnt signaling pathways, as well as cell adhesion molecules, NOD-like receptor signaling pathway. Notably, we noticed that tumor-like EPC were highly expressed genes related to H. pylori infection (Figure 2I). Taken together, these data revealed the diversity and distinct cell states of epithelial cells in the cascade of gastric carcinogenesis. Remarkably, these findings further revealed the oncogenic transcriptomic features and biomarkers (such as KRT17), functional characteristics, and cell origin of tumor-like EPC subpopulation that were abundant in GC tissues, and finally identified gene sets related with H. pylori infection.

A distinctive epithelial cell population enterocyte from precancerous lesion gastrointestinal metaplasia

Given that the enterocyte cell populations were overwhelmingly enriched in gastrointestinal tissues, as determined by sample-to-sample breakdown of proportional cell cluster (Figure 2C), we aimed to characterize the transcriptional patterns and enriched pathways of this distinct cell cluster. The CDX family members CDX1 and CDX2, which have been previously reported to confer gastrointestinal metaplasia phenotypes(20), were mainly expressed in enterocyte cells (Figure S3A). This further confirmed the specificity of enterocyte subclusters in gastrointestinal metaplasia, the important precursor lesion of GC. Spasmolytic polypeptide-expressing metaplasia (SPEM) is a regenerative pathological state of gastric corpus mucosa in response to injury, associated with loss of chief and parietal cells, and hyperplasia of neck mucous cells(21). We observed that SPEM cells, defined by the co-expression of MUC6 and TFF2, seemed to be gradually decreased from GS to IM, and GC (Figure S3B). This is consistent with the notion that SPEM represents as a precursor lesion to metaplasia(17, 18). To assess the transcriptional features of enterocyte cells, we compared the DEGs of this subcluster with other epithelial cell subsets. The UMAP plot showed that these marker genes that regulate lipid metabolism (FABP1, APOC3 and APOC4), and specific genes for intestine (KRT20, ANPEP) were specifically overexpressed in enterocytes (Figure 3A). Based on the pseudotime results using Monocle 3 algorithm, these candidate marker genes were significantly upregulated in metaplasia lesion (Figure S3C). We further performed IF staining to delineate the existence of enterocytes expressing marker genes FABP1 and KRT20 (Figure 3B). As expected, FABP1 and KRT20 were remarkedly overexpressed in IM tissues, compared with GS or GC tissues by IHC staining (Figure 3C).

Gene Ontology (GO) enrichment analysis indicated that enterocytes expressed genes were involved in chylomicron formation and cholesterol homeostasis, suggesting the activity of energy metabolism in IM phase. Enriched Signaling pathways were involved in fat digestion and absorption, cholesterol metabolism and PPAR signaling pathways (Figure S3D). Afterwards, we applied SCENIC to infer TF-target regulatory network (regulons) of enterocytes. HNF4G, exhibited specific and high transcriptional prosperities in enterocytes (Figure 3D), which appeared to be consistent with the previous studies of the important role of HNF4G in regulation of fatty acid oxidation and intestinal stem cells renewal(22). Of note, HNF4G, together with its downstream 83 target genes, such as FABP1, FABP2, APOB, and VIL, were specifically expressed in enterocytes (Figure 3E and Table S3). IHC staining verified the increased levels of HNF4G in IM samples when compared with GS or GC samples (Figure 3F and G). In summary, these findings uncovered enterocytes as the distinct epithelial sublineages in premalignant intestinal metaplastic lesion and HNF4G as its specific transcriptional regulator.

Differentially expressed genes associated with H. pylori infection in epithelial cells

Given that chronic infection with H. pylori causes gastric mucosal injury and initiates neoplastic transformation, we aimed to identify the DEGs associated with H. pylori in epithelial cells. H. pylori preferred to colonize the pit cells of gastric epithelial mucous layer, as indicated by IF staining of MUC5AC and H. pylori (Figure 4A), consistent with previous studies(23). By comparing the cell number of epithelial subsets, endocrine cells expressing GAST and SST were reduced in H. pylori -positive samples. Conversely, as previously observed enrichment in GC, enterocyte and tumor-like EPC subclusters were increased in H. pylori-positive tissues. These data also implied the close association significant differences in the regulation of different cell types by H. pylori in gastric carcinogenesis (Figure 4B). Differential gene expression analysis showed H. pylori -positive epithelial cells had relatively lower expression levels of genes involved in gastric acid and digestive enzymes secretion (e.g., PGC, LIPF, and GAST), but higher levels of genes that specifically overexpressed in metaplasia lesions (PHGR1, FABP1, ANPEP, KRT20, PRAP1 and DMBT1) and extracellular matrix OLFM4 oncogene (Figure 4C; Figure S3E and Table S4). In particular, OLFM4, a stem cell marker that could regulate cell adhesion(24), was further validated to be overexpressed in GC tissues compared to GS and IM tissues by RT-PCR and IHC staining (Figure 4D and Figure S3F), as consistent with previous reports(25). Notably, H. pylori -positive GC specimens exhibited higher levels of OLFM4 than H. pylori-negative subjects (Figure 4E). Public datasets analysis also confirmed the upregulation of OLFM4 in human stomach carcinoma (Figure S3G). In accordance with scRNA-seq, ANPEP, PRAP1, FABP1 and KRT20 were significantly augmented in IM tissues. In particular, they were expressed at higher levels in H. pylori-positive IM individuals, although without statistically significant differences (Figure 4F-I), indicating their potential as diagnostic markers for H. pylori -associated IM lesion.

H. pylori infection induced intestinal metaplasia in INS-GAS mice and initiated enterocyte features.

To determine the effect of H. pylori infection on the characteristics of enterocytes, transgenic INS-GAS mice which overexpress human gastrin and can develop gastric metaplasia and dysplasia with H. pylori infection were infected with H. pylori PMSS1 strain for 4 months. H. pylori colonization was confirmed by Warthin-Starry silver staining (Figure S3H). As shown in Figure 5A, PAS and H&E staining showed that the gastric mucosa of infected INS-GAS mice developed intestinal metaplasia lesion. Also, H. pylori infection resulted in more severe gastric epithelial defects and gland atrophy(26) (Figure 5B). IHC staining was further performed to detect the expression of FABP1 and KRT20, which are the marker genes for enterocytes subtype. As expected, there were significant increases in the levels of FABP1 and KRT20 in gastric mucosa with H. pylori infection, compared to uninfected mice (Figure 5C and D). In addition, multicolor immunostaining confirmed that FABP1 and KRT20 expression were significantly upregulated in the gastric mucosa of H. pylori infected mice (Figure 5E). Since we showed that HNF4G as a specific transcription factor for IM precancerous lesion, we further examined the expression in the gastric mucosa of infected and uninfected mice. As expected, H. pylori infection was observed to remarkably induce HNF4G expression (Figure 5F). Taken together, these in vivo data indicated that H. pylori infection could induce a gastric enterocyte cell feature, such as the upregulation of FABP1 and KRT20, and enhancement of the transcription factor HNF4G, which may be involved in the development of precancerous lesion of IM.

The heterogeneity of ECs and ECs-tumor-like EPC subtypes interactome predictions

ECs play a crucial role in tumorigenesis by governing angiogenesis to provide nutrients and oxygen for cell proliferation (27). Given the gradual increase in ECs proportion in gastric carcinogenesis cascade, we next aimed to detailly investigate the transcriptomic heterogeneity of ECs and to identify tissue-specific subpopulations. We obtained 6266 high-quality ECs in which unsupervised clustering revealed 7 distinct subclusters (Figure S4A). We identified 4 cell subtypes belonging to traditional vascular beds, i.e., venous ECs (VCAM1, SELP, ACKR1), lymphatic ECs (LYVE1, PROX1), capillary ECs (PLPP3, RGCC, CA4) and arteries ECs (SEMA3G, GJA4, GJA5) (Figure 6A and Figure S4B). A cluster of subpopulations was designated as tumor-like ECs due to remarkably increase in GC tissues, which highly expressed marker genes known to be enriched in angiogenesis and extracellular matrix (ECM) remodeling (LGALS1, COL4A1, COL4A2 and ESM1), as well as canonical smooth muscle cells marker (TAGLN) (Figure 6B-D). Another population Endo_3 (cluster 3) exhibited significantly lower UMI counts compared to others, with highly expressed lncRNAs (e.g., HES1, MALAT1, AC020916.1), was not be further analyzed in depth (Figure S4C). Overall, we observed that all EC subsets were largely enriched in GC tissues, compared with those in GS and IM tissues. Both venous and capillary populations accounted for higher percentage of all ECs, whereas lymphatic population were rare (Figure 6B). However, these populations manifested no significance between H. pylori -positive and -negative patients, except for tumor-like ECs. Tumor-like ECs were more abundant in GC samples, especially in H. pylori-positive samples (Figure S4D and E). Additionally, enriched pathways of the highly expressed genes in tumor-like ECs were involved in cell adhesion and PI3K/Akt signaling pathway (Figure 6E and F). Of note, genes enriched in these two pathways were greatly upregulated in GC tissues, especially in H. pylori -positive subgroup, suggesting the role of H. pylori infection in promoting angiogenesis (Figure 6G). Since GC originated from the abnormal changes in epithelial cells(28), and tumor-like EPC was the specific epithelial-derived subset in GC, we speculated that ECs and tumor-like EPC cluster communicate closely. Using CellChat(29), we found that endothelial subtypes dominantly drive outgoing signaling, while tumor-like EPC was the target for incoming signaling (Figure 6H). In detail, the COLLAGEN and VEGF signaling pathways were highly activated between their communications (Figure S4F and G). COL4A1 and COL4A2 were the most highly expressed ligand by endothelial subclusters, especially by tumor-like ECs, and CD44 and SDC44 as their receptors on the tumor-like EPC. In turn, the pro-angiogenic VEGF signaling pathway was mainly driven by VEGFA ligand from tumor-like EPC to VEGFR receptors (FLT1, KDR) from ECs (Figure 6I and J). We also independently revealed that COL4A1 and COL4A2, the secreted factors of COLLAGEN signaling pathways, were expressed at high levels in gastric malignancies compared with paired normal tissues by reanalyzing bulk RNA-seq gastric carcinoma GEO datasets (GSE 122401, GSE65801, GSE13195) and TCGA database (Figure S5A and B). However, only some RNA-seq data from public databases showed that the receptors SDC4 and CD44 were upregulated in gastric carcinomas (Figure S5C and D). Altogether, these findings highlighted the molecular signatures of GC tissue-specific EC subsets, and further predicted ECs and tumor-like EPC interactions that may contribute to gastric carcinogenesis microenvironment.

Distinct fibroblast cell subtypes in cancerous and non-cancerous samples

Fibroblasts are highly organized stromal cells embedded in the ECM. Accumulating evidence indicated high diversity of fibroblasts, which exhibited phenotypes and functions associated with specific-cell lineage(30, 31). To explore fibroblasts cell lineage states, we defined 11 subsets using unsupervised clustering method. Cluster 11 were mixed cells and was disregarded of further analysis (Figure S6A). Surprisingly, compared to non-cancerous tissues, we observed that fibroblasts in gastric carcinoma tissues exhibited completely different lineage states, hence named cancer-associated fibroblasts (CAFs) (Figure 7A). In detail, ECM fibroblasts (ECM_Fib; expressing COL5A1 and POSTN) and myofibroblasts (ACTG2) were characterized, mainly derived from GS and IM tissues. Swann cells was named due to specifically overexpression of S100B and PLP1(32). There were three CAFs subpopulations, consisting of antigen-presenting CAFs (apCAFs; expressing HLA-DRA and HLA-DRB1), epithelial-mesenchymal transition CAFs (EMT_CAFs; expressing KRT8 and KRT19), and inflammatory CAFs (iCAFs; expressing FBLN1, IGFBP6 and CXCL1) (Figure 7B; Figure S6B and C)(33). To further explore the gene set signature of CAFs, we next performed a gene set enrichment analysis (GSEA) and differential gene expression analysis by comparing gene functions of all CAFs (iCAFs, apCAFs and EMT_CAFs) versus other fibroblast subsets (ECM_Fib and Myofibroblasts) (Figure 7C and D). Of note, CAFs were enriched for gene sets involved in immune and inflammation-related pathways (IGFBP4, CXCL8, CXCL2), complement cascades (C3, C7), and Wnt signaling pathway (SFRP2, SFRP4), but exhibited lower expression of gene sets involved in metabolism (Figure 7C and D; Table S5). These results suggested the crucial role of CAFs in shaping tumor immune microenvironment.

Given the higher fraction of iCAFs in GC tissues (Figure 7E), we explored the characteristics of this subpopulation We observed that iCAFs displayed high expression levels of complement components (C3 and C7), extracellular regulators of Wnt signaling (SFRP2 and SFRP4)(34). Also, iCAFs showed largely increased expression of PLA2G2A, like the previously reported immune-infiltrating PLA2G2A⁺CAFs in breast cancer(35) (Figure S6C). Dual-color IF staining validated the abundance of iCAFs (COL1A1⁺FBLN1⁺ or COL1A1⁺C3⁺) in GC tissues (Figure 7F). We next investigated the dynamic transition of fibroblast cells. In pseudotime analysis, we removed schwann cells, due to their completely different characteristics. As expected, this analysis indicated that ECM_Fib and myofibroblasts were at the beginning of the trajectory path, whereas all kinds of CAFs were at terminally differentiated states (Figure S6D). Of note, iCAFs appeared to arise from ECM_Fib cell clusters (Figure 7G). Moreover, according to the regulatory network using SCENIC analysis, NFATC1-driven regulons and transcriptional factor expression were inferred to be specific and enriched in iCAF subsets (Figure S6E and F).

Fibroblast-tumor-like EPC populations interactome prediction and identification of DEGs in fibroblast cells associated with H. pylori

Given the increased attention of fibroblasts in multicellular interaction in tumor microenvironment(36), we next predicted the interaction between fibroblasts and tumor-like EPC, to explore the mechanism through which microenvironment favors gastric tumorigenesis. Interestingly, CellChat analysis revealed that fibroblast subsets, especially iCAFs, dominantly drive outgoing signaling, which means secreting more ligands. The corresponding incoming receptors were expressed in tumor-like EPC (Figure 7H). In agreement with the analysis of ECs-tumor-like EPC interaction, the COLLAGEN signaling members, such as COL1A1 and COL1A2, as well as Fibronectin 1(FN1) signaling, were highly ranked ligands expressed by fibroblasts. Their corresponding receptors in tumor-like EPC were similar, such as CD44 and SDC4. In addition, the APP/CD74 and FGF7/FGFR2 also contributed to fibroblasts-tumor-like EPC interactions (Figure 7I).

Existing evidence indicates the critical role of H. pylori on gastric epithelial cells or macrophages, stromal cells response to H. pylori remains unclear. Given that the high heterogeneity of fibroblast cells in gastric malignant transformation, we analyzed the effect of H. pylori on fibroblast cell composition. Of note, H. pylori -positive gastric specimens showed increased percentage of CAFs, as indicated by apCAFs, EMT_CAFs and iCAFs, whereas relatively reduced ECM_Fibs (Figure 7J). Identification of DEGs between H. pylori-positive and -negative groups showed that several ligands and receptors for cellular signaling pathways were the top DEGs, such as FN1, C3, IL1B, CXCL8 and CD74 (Figure 7K and Table S6). They were also significantly overexpressed at bulk RNA levels in GC tissues, particularly in H. pylori -positive GC tissues (Figure S7A). We further queried bulk RNA-seq datasets from the TCGA and GEO databases, which also showed higher expression of FN1, CXCL8, C3, but not IL1B, in gastric cancerous tissues than that in paired adjacent normal tissues (Figure S7B-E). Consistently, we performed RT-PCR analysis in an expanded sample size to confirm the elevated mRNA levels of FN1 and CXCL8 in GC tissues, particularly those associated H. pylori infection (Figure 7L and M). Altogether, these findings suggest that H. pylori infection may drive the interaction between fibroblasts and epithelial cells in GC microenvironment via activation of FN1 signaling network.

Tissue-specific patterns of myeloid cells in H. pylori -related gastric carcinogenesis

To characterize the subsets of myeloid cells that showed markedly increase in GC, we identified five common cell subsets based on canonical cell markers, consisting of macrophage (CD68, CD163, MRC1), Monocyte (VCAN, LYZ), type 1 conventional dendritic cells (cDC1; expressing XCR1 and CLEC9A), type 2 cDC (cDC2; expressing CD1C and CLEC9A), and mature dendritic cells (mDC, expressing LAMP3) (Figure 8A and Figure S8A). A specific cluster of TAM was designated, based on highly expressed marker genes of macrophages and enriched in tumor tissues (Figure 8A and B). Cluster 4 were low quality cells with low UMI contents, and cluster 8 were mixed cells, both of which were disregarded of further analysis (Figure S8B). Of note, total myeloid cells number were much higher in GC tissues, especially H. pylori -infected GC tissues, than that in GS and IM tissues (Figure 8B and Figure S8C). Except for TAM, monocytes also showed a relatively high proportion in GC, whereas the cDC2 subtype tended to gradually decrease in gastric cancerous cascade (Figure S8D). The complement components (C1QA, C1QB and C1QC), and cholesterol metabolism genes (APOE and APOC1) were enriched in macrophages, as described previously(37). Monocytes highly expressed a set of neutrophil-associated genes (S100A8, S100A9), pro-inflammatory cytokine IL1B and inflammasome NLRP3 (Figure 8C). Dual-color IF staining further validated the presence of TAM subsets, as indicated by co-expression of CD68 and CXCL3. Likewise, monocytes were verified by combined immunostaining of CD68 and S100A9 (Figure 8D). GSVA analysis revealed enrichment of monocyte gene sets involved in metabolism pathways, such as fatty acid biosynthesis, bile secretion. The gene sets of TAMs were involved in IL-17, TNF pathway and cytokines interactions (Figure S8E). To clarify M1 and M2 polarization of TAMs, we calculated pro-and anti-inflammatory scores using related gene sets (Table S7). In agreement with previous studies(15), TAMs exhibited dominant M2-like phenotype, suggesting its immunosuppressive role in tumor microenvironment (Figure S8F). Pseudotime analysis using monocle 2 suggested that both TAM and monocyte subsets in neoplastic tissues were likely derived from macrophages in GS and IM tissues (Figure 8E).

The interactome network between myeloid and epithelial cells was further studied. Interestingly, the outgoing and incoming signaling patterns were obviously observed between myeloid cell subpopulations in tumor microenvironment, compared to myeloid-epithelial cells. Macrophages, TAMs, monocytes and cDC2 cells not only secreted cytokines, but also expressing corresponding receptors (Figure S8G). To identify myeloid cell-derived genes regulated by H. pylori, we compared the DEGs from myeloid cells between H. pylori -positive and H. pylori -negative tissues. Notably, myeloid cells from H. pylori -positive tissues displayed high expression of CXCL3, TIMP1, CD14 and CTSL genes (Figure 8F and Table S8). Further analysis of gastric carcinoma bulk RNA-seq datasets from GEO and TCGA database showed that CD14, CTSL and TIMP1 were significantly upregulated in GC tissues compared to the paired adjacent non-cancerous tissues (Figure 8G). Of note, H. pylori -positive patients showed even higher expression (Figure 8H and Figure S9A-C). Likewise, these results were further validated by qRT-PCR analysis with additional samples (Figure 8I).

T, B and plasma cell clustering and subtype analysis

We next applied unsupervised clustering of immune T and B cells to reveal their intrinsic substructures. A total of 14 subclusters of T cells were classified based on well-known markers CD4 and CD8 (Figure 9A). The first CD4 cluster CD4_IL7R highly expressed Naïve marker IL17R that is critical for T cells activation and development(38). The second cluster CD4_PDCD1, showing gradually reduced proportion in gastric carcinogenesis cascade, specifically expressed T follicular helper markers, such as TOX2 and PDCD1 (39, 40). The final cluster CD4_FOXP3 was well-known Treg cells with immunosuppressive effects based on the dominant CTLA4, FOXP3 expression(41) (Figure 9A-C). According to signature genes expression and top-ranked genes, CD8⁺T cells comprised 7 subtypes: CD8_GZMK and CD8_GNLY cells highly expressed cytotoxicity-associated genes (GZMK, GZMB). CD8_IFNG showed activated resident memory-related markers (CCL4, IFNG, CCL3)(42, 43); CD8_CAPG exhibited elevated expression of CAPG and novel immune checkpoint KLRC1(44). CD8_IL17A (IL17A, KLRB1) cells showed similar activation mode and cytokines secretion with TH17 cells(45). CD8_XCL1 cells showed enriched expression of TRDC, XCL1, XCL2, and CD160 genes. The CD8_MKI67 cluster was designated based on uniquely expression of MKI67(40, 46) (Figure 9A, B and D). In addition, three clusters were characterized as DNT (HSPA1A, HSPA1B), C14_KRT81 (KRT81) (47)and NK cells (NKG7).

We found that Treg cells represent higher proportion in GC samples than that in GS or IM sample (Figure 9E-F). When assessing the role of H. pylori on T cell heterogeneity, we observed a trend of increased CD8_IL17A subset in H. pylori -positive (versus H. pylori -negative) GS tissues, consistent with previous reported viewpoints regarding the role of IL17 cytokine in driving protective immunity against pathogen and promoting inflammatory response(45) (Figure 9F). Combined IF staining labeled CD4 and FOXP3 also revealed Treg cells to be enriched for GC, as previously reports where Treg cells are increased in several solid malignancies and associated with poor prognosis(48, 49) (Figure 9G). The trajectory analysis using monocle 3 indicated that CD4⁺T cells originated from CD4_IL7R cells and then differentiated into CD4_PDCD1 and CD4_Treg cells (Figure 9H), as previously described that the sustained expression of IL7R in the naïve states is critical for the maintenance and development of all T subpopulations(38). As for CD8⁺T cells, we observed a notable enrichment of the CD8_MKI67 subpopulation in GC tissues (Figure 9F). Multi-color IHC staining for CD3, CD8 and Ki67 further confirmed the presence of CD8_MKI67 cells in gastric carcinoma tissues (Figure 9I), supporting the previous findings indicating the proliferative function of T cells in tumors(42). Taken together, these findings suggested the potential functional role of Treg cells, and IL17A-expressing and cytotoxic characteristics CD8⁺T cells in H. pylori -associated gastric neoplasm.

The re-clustering of B and plasma cells revealed 6 populations, including active B (CD69), Naïve B (TCL1A, IL4R), Pre-pro B (CD7, IL7R), IgA plasma (IGHA1, IGHA2), IgG plasma (IGHG1, IGHG2), and IgM plasma cells (IGHM) (Figure S10A and B). Overall, B and plasma cells defined clearly separate clusters, with the higher abundance in GC samples (Figure S10C). The relative cellular proportion of plasma cells various across tissues. IgA⁺ plasma cells were highly prevalent in GS/IM, while IgG⁺ plasma subsets were enriched in GC samples (Figure S10D), in agreement with previous work(50). Only slight difference was existed between H. pylori -positive and negative gastric tissues (Figure S10E). Nevertheless, IgG⁺ plasma population appears to be more abundant in H. pylori -infected individuals (Figure S10F).

Multicellular communication network shapes H. pylori -driven carcinogenesis microenvironment

To understand the cellular signaling pathways driving gastric carcinogenesis cascade, we used CellChat to infer autocrine and paracrine signaling network among all cell types, including epithelial cells, stromal cells and immune cells. The analysis suggested that stromal cells were the dominant senders of signaling, while immune T and myeloid cells were the receivers (Figure 10A). ECs, as well as fibroblasts, also showed relatively strong interaction with epithelial cell subpopulations, pit mucous cells and tumor-like EPC (Figure 10B and Figure S11A). The COLLAGEN signaling component, which included the classic marker COL1A1, was mainly driven by fibroblasts and almost implicated in communication with all other cell types (Figure S11B). Another fibroblast-derived C3 signaling pathway uniquely recruited to myeloid cells through C3-C3AR1 ligand-receptor pair (Figure S11C and D). In line with the previous results, the VEGF signaling pathway was dominant in ECs and tumor-like EPC interaction, as well as ECs-myeloid cells interaction (Figure S11E). Additionally, we found that epithelial-derived pit mucous cells secreted IL1B and that recruited ILR2-expressing myeloid cells (Figure S11D and F). Importantly, the communication network of all cell subtypes inferred that iCAFs subsets with high expression in GC dominantly drive outgoing signaling (Figure 10C). Due to the significant increase of myeloid cells in GC and the immunomodulatory role of iCAFs, we further focused on exploring the interaction between iCAFs and myeloid cell subsets. The analysis suggested that iCAFs primarily secreted cytokines and were recruited to myeloid cells, particularly the TAM subpopulations (Figure 10D). The COL1A1, MIF, C3, and APP signaling regulatory networks were mainly responsible for the communication between iCAFs and myeloid cells (Figure 10E). Based on the exclusive enrichment of iCAFs and myeloid cells, these findings suggest their interaction as an important decisive factor for poor prognosis in GC.

We then compared the overall communication probability in H. pylori positive and negative samples. Notably, H. pylori -infected gastric tissues exhibited a significant increase in the number and strength of inferred intercellular communications (Figure 10F). Moreover, our ligand-receptor pairs analysis predicted several enriched pathways in H. pylori -positive specimens, such as SPP1, TNF, THY1 and complement pathways, while GAS and IL10 signaling networks were more abundant in H. pylori -negative specimens (Figure 10G). We subsequently focused on TNF pathways that has been implicated in inflammatory diseases(51). The analysis predicted TNF-TNFRSF1A (TNFR1) was found to be highly active in H. pylori -positive samples, suggesting the important role of TNF-TNFR1 pathway in the pathogenesis of H. pylori infection (Figure 10H and I). These results highlighted the putative importance of intercellular crosstalk in H. pylori -induced gastric neoplastic transformation.

Gastric adenocarcinoma has the high incidence and death rates, especially in Eastern Asia countries. Non-cardia cancer, the most common GC and arises in distal regions of the stomach, is almost attributable to H. pylori infection(52). Non-cardia GC is most commonly of the intestinal-type gastric adenocarcinoma, which develops through a sequential progression of gastric lesions known as the Correa cascade(3). In this study, we applied scRNA-seq to comprehensively analyze the cellular transcriptomic heterogeneity and communication among all cell types including epithelial, stromal and immune cells in gastric carcinogenesis microenvironment, and characterized specific cell populations linked to gastrointestinal metaplasia and gastric carcinoma (Figure S12A). Importantly, our study further discovered the role of H. pylori infection in different cell subtypes across GC progression and identified potential biomarkers for H. pylori -associated GC (Figure S12A and B). Our findings were orthogonally validated by expanding sample size and utilizing bulk RNA-seq data from public GEO database and TCGA-STAD cohort, and mechanically confirmed in transgenic INS-GAS mice that has been widely used for studying the effect of H. pylori infection. Based on our findings, we proposed a working model to provide a novel and comprehensive landscape of H. pylori -associated-gastric carcinogenesis ecosystem (Figure S12).

Previous studies have explored the diversity of cell types in gastric precancerous and cancerous lesions(53, 54), but the analysis was limited to specific cell populations within small sample size. For example, kumar et al. compared the tumor microenvironment of gastric cancer tissues with different Lauren pathological subtypes, with specific focus on investigating the states and transcriptomic patterns of plasma cells and fibroblasts lineages(9). Zhang and colleagues performed scRNA-seq in human gastric tissues (56440 cells) to study the molecular features of epithelial cells in gastric precancerous lesions and early GC(55). This study innovatively lies in scRNA-seq studies on uniform samples of progressively gastric pathological lesions (6 of GS, 6 of IM, and 6 of GC), and obtained a larger cell number size (༞90000 cells), to create a cellular heterogeneity landscape of all cell subtypes during gastric carcinogenic cascade. A recent single-cell study also comprehensively investigated the transcriptomic heterogeneity of stromal and immune cells. There were several concordant findings with our study, such as high enrichment of IgA⁺plasma cells in precancerous lesions, stromal cells remodeling and high abundance of M2-like TAM cluster in GC(50). Strikingly, our study further emphasized the interaction among all cell subtypes and determined the role of H. pylori infection in the heterogeneity and expression profiles of different cell types across gastric pathological stages.

Gastric adenocarcinoma often preceded from precancerous lesions, such as metaplasia and dysplasia, which originate from alteration in epithelial cells (56). Epithelial cells contain various cell populations based on molecular profiling, making it challenging to understand the molecular biological changes of each cell type in gastric neoplastic lesions until the emerging of single-cell sequencing technology. Zhang and colleagues performed scRNA-seq study to delineate cellular and differentiation heterogeneity of epithelial-derived malignant cells in gastric adenocarcinoma(10). Here, we focused on the dynamic changes of epithelial cells during gastric carcinogenic cascade. Our data indicated that there was a gradual decrease in some subpopulations of epithelial subtype in gastric tumorigenesis, including pit mucous cells, neck mucous cells, chief cells, parietal cells and all endocrine cells, which were responsible for maintain functions of epithelial cells by secreting acid, digestive enzymes, mucus and gastrin. This also aligns with the widely accepted notion that the disruption of epithelial homeostasis marks the initiation of gastric mucosal injury(28). Surprisingly, we observed the enrichment of two distinct epithelial cell subpopulations, i.e., enterocytes in IM tissues and tumor-like EPC in GC tissues, especially in H. pylori -infected samples. Enterocytes exhibited specific high expression of intestinal mucosa markers FABP1, ANPEP and KRT20, which were enriched in metabolic pathways. This finding is consistent with a recent study identifying the gastric epithelial subsets with similar characteristics(11). To identify the transcription factor for enterocytes, we determined HNF4G that has been previously shown to regulate intestinal differentiation(22, 57), suggesting its crucial role in gastrointestinal metaplasia. In addition, there was an extremely abundant subset of epithelial cells in GC tissues, defined as tumor-like EPC with the following features: 1. Higher CNV numbers; 2. Significant high expression of KRT17 and CEACAM5; 3. Their origin from pit mucous cells. Notably, the gene set of tumor cells was involved in H. pylori infection-associated pathways, supporting the unequivocal role of H. pylori in gastric tumorigenesis. The potential oncogene OLFM4, as well as signature genes of enterocytes (ANPEP, PRAP1, FABP1, KRT20), could be regulated following H. pylori infection, which was validated by RT-PCR analysis of our further expanded samples. Earlier report showed two specific subpopulations in IM, one marked by OLFM4 responsible for intestinal stem cell phenotype and the other HES6-marked globet cells(58). To some extent, this seems to contradict our findings, and one possible explanation might be due to the difference in IM sample size.

Fibroblasts were highly organized constituents of stromal cells. Functionally, fibroblasts are fundamental regulators of ECM synthesis and supporting nearby epithelial cells growth and differentiation by paracrine signaling(30, 59). Recently, CAFs, widely considered as the distinct cell population in tumors, have been shown to play cardinal roles in shaping tumor environment. Accumulating evidence indicates that CAFs largely influence tumor initiation, growth and dissemination through ECM remodeling, tumor immunity and angiogenesis(60). In this study, compared to GS and IM, gastric carcinoma presented a completely different fibroblast cell states, named CAFs. Immunomodulatory pathways were significantly upregulated in CAFs, such as antigen presentation, Nod-like receptor and Toll-like receptor pathways. In detail, we observed that CAFs in GC exhibited high heterogeneity, including apCAFs, iCAFs and EMT_CAFs. These results were in line with the earlier studies that have also identified apCAFs and iCAFs, whose function in regulating tumor immunity may be responsible for tumor immune evasion(61, 62). In addition, given that H. pylori favors epithelial cell colonization and causes inflammatory response, previous studies have mostly focused on epithelial cells or macrophages, instead of exploring the effects of H. pylori on stromal cells. Here, we compared the cellular heterogeneity and expression programs in fibroblast subpopulations between H. pylori infected and uninfected tissues. Bulk-RNA seq data from GEO and TCGA cohort and experiments with expanded gastric samples, have shown that fibroblast-secreted ligands FN1 and CXCL8 may be involved in the pathogenesis of H. pylori. In fact, CXCL8, widely expressed in various cell types, has been well known to be crucial inflammatory chemokine induced by H. pylori. Nevertheless, the molecular mechanisms underlying the pathogenesis of H. pylori infection remain to be further determined.

Myeloid cells, constituting a key cellular component of immune cells, plays an essential role in the pathogenesis of bacterial infection. Macrophage is the major type of myeloid cells that could efficient phagocytosis and kill H. pylori (63). However, the persistent infection with H. pylori exerts inflammation and subsequent carcinogenesis through influencing macrophage function, implying the substantial heterogeneity of macrophages in H. pylori -associated gastric lesions. In agreement with the recent studies that revealed a remarkable enrichment of myeloid cells in several solid tumors(64, 65), we also confirmed the significant role of myeloid cells in gastric carcinoma. Additionally, we identified distinct subsets of TAMs and monocytes in GC tissues, and revealed their enriched immunometabolic functions and inflammatory pathways, suggesting the potential mechanisms of tumor immunosuppression. Moreover, our data provides evidence supporting that the involvement of CD14, CTSL and TIMP1 genes expressed in myeloid cells in H. pylori -induced carcinogenesis. Further studies are warranted to reveal their functions and the underlying molecular mechanisms.

The complex interactions between epithelial cells, stromal cells and immune cells shape the tumor ecosystem and profoundly modulate tumor biology. Accumulating evidence indicates the crucial role of interactions among immune cells feature immunosuppressive and pro-tumorigenic functions(66). The intercellular communication network between all cell types in gastric carcinogenesis. This study highlighted the pivotal role of stromal cells, particularly iCAFs and tumor-like EC subsets, in serving as key secretory ligands, while immune cells express receptors. The initiation of gastric carcinogenesis is closely linked to alterations in epithelial cells, however, the communication between epithelial cells and other cell types with tumors remains poorly understood. We analyzed the crosstalk between tumor-like EPC and stromal cell populations. In detail, the COLLAGEN signaling network was indicated to mediate the crosstalk between stromal cells and tumor-like EPC. Additionally, the VEGF signaling was involved in the tumor-like EPC communication with ECs, facilitated by ligand-receptor pairs VEGFA and FLT1, KDR. Notably, our findings revealed that the intercellular communications were more active in H. pylori -positive gastric tissues. Several significant signaling pathways, such as SPP1, TNF and THY1 signaling networks, were indicated to regulate the cellular microenvironment in response to H. pylori infection. Therefore, to prevent H. pylori -related gastric mucosal lesions, it is crucial to consider not only the factors related to individual cell types but also the complex communications between various cell types. Documented evidence shows that the interaction between cancer-derived fibroblasts and metaplastic epithelial cells promotes transition of SPEM-expressing metaplasia into dysplasia lesions(67). Future studies should consider how intercellular interactions contribute to GC initiation and progression, which would provide novel insights for the diagnosis and treatment of GC. In-depth mechanistic studies of interaction between ligands and receptors in the tumor microenvironment would pave the way to identify novel therapeutic targets in GC treatment.

In summary, as one of the largest single-cell analysis of Correa’s cascade of gastric carcinogenesis to date, we comprehensively unveiled both common and specific characteristics of nearly all cell types in gastric precancerous and cancerous lesions. Notably, for the first time, we provide insights into the influence of H. pylori infection on cellular heterogeneity in the gastric microenvironment, which may contribute to a better understanding of the pathogenesis of H. pylori, and further offer valuable resources for the prevention and early diagnosis of GC. Furthermore, our data support the need for further in-depth research focusing on epithelial cells biology in the context of H. pylori -induced gastric carcinogenesis, such as the regulatory mechanisms of enterocyte subsets for gastrointestinal metaplasia, the interactions between epithelial cells and other cell types, through in vivo and in vitro experiments. The role of CAFs, tumor-like ECs and TAMs in gastric carcinoma microenvironment also carries the potential as therapeutic targets and predictive biomarkers.

Human tissue samples

A total of 18 patients pathologically diagnosed with gastric lesions at The First Affiliated Hospital of Nanchang University, including 6 of GS, 6 of IM, and 6 of GC, were enrolled in this study. The specimens of GS and IM were obtained during endoscopy, and the GC specimens were obtained from surgical samples without adjuvant therapy. The patients with gastric mucosal lesions were then divided into two groups according to H. pylori status. The sections with H&E staining were reviewed by senior pathologists. Status of H. pylori infection for these clinical specimens was determined by ¹³C-urea breath test (UBT), immunohistochemistry (IHC) and immunofluorescence (IF) staining. Additionally, formalin-fixed paraffin-embedded (FFPE) specimens including GS, IM and GC were collected from 120 patients for experimental validation. The clinical characteristics of patients were summarized in Table S1. This study protocol was approved by the Ethics Committee of the First Affiliated Hospital of Nanchang University under protocol number of (2023) CDYFYYLK (01-009). All tissue samples were obtained under informed consent.

Tissue dissociation and single-cell suspension preparation

The tissue samples were removed from patients and quickly immersed in Dulbecco’s modified eagle medium (DMEM; Cat#11995065, Gibco) containing 10% fetal bovine serum (FBS; Cat#10091148, Gibco). Samples were minced into about 0.5 mm³ pieces and enzymatically digested in DMEM supplemented with trypsin (Cat#15090046, Gibco), collagenase I (Cat# 17018029, Gibco), collagenase II (Cat# 17101015, Gibco) and collagenase IV (Cat# 17104019, Gibco). The samples were incubated and digested at 37℃ for approximately 60 min and then filtered using a 100 μm strainer (Corning). Following centrifugation (300 g) at 4℃ for 5 min, the cells were resuspended with medium containing red cell lysate (Cat#130094183, MACS). After standing 10min at 4℃, the single-cell suspension was centrifuged at 300 g for 5 min and resuspended in 100 μl medium. Cell concentration and viability were detected under Luna Cell Counter. Detailed description of single-cell sequencing and data analysis were provided in online supplemental methods.

Single cell droplet-based RNA sequencing

The single-cell suspensions were adjusted to 700-1200 cell/μl. The suspensions were then converted to separate barcoded scRNA-seq libraries using the Chromium Next GEM Single Cell 3ʹ Reagent Kits v3.1 (Cat#1000268, 10×genomics) according to the manufacturer’s instruction. Libraries were sequenced on an Illumina Nova 6000 PE150.

Single-cell RNA-seq data processing

The raw reads were processed to gene expression matrices using Cell Ranger (10×genomics). The Seurat software package was used for further quality control. To remove low quality cells and likely multiple captures, the following quality control steps: the cells with gene numbers less than 200, UMI less than 1000 and log10Genes Per UMI less than 0.7 were filtered out. Furthermore, the cells where ＞20% of the counts belonged to mitochondrial genes and 5% of the counts belonged to hemoglobin genes were removed. Additionally, the DoubletFinder package (version 2.0.2) was used to identify potential doublet. Finally, 124528 cells (Median: 5319 cells/patient) were obtained for the downstream analysis. We obtained 1717 genes and 5763 UMIs per cell on average. The detailed sequencing data of quality control were shown in Table S1.

Cell-type clustering and marker identification

The FindVariableGenes function from Seurat R package was used to screen highly variable genes (HVGs). Then the principal component analysis (PCA) was performed based on HVGs, followed by visualization using Uniform Manifold Approximation and Projection (UMAP).

To detect cluster-specific expressed genes, the FindAllMarkers from Seurat R package was used to compare the differential genes of a given cluster with all other clusters.

CNV analysis

To investigate large-scale chromosomal CNV, the ScRNA-seq data obtained from these tissues were processed using InferCNV, a robust computation tool specifically sesigned to infer CNV numbers within individual cells based on intercellular variability in gene expression patterns. In this study, we employed InferCNV to explore CNV numbers in epithelial cells subsets with H. pylori infected or uninfected gastric tissues.

Pathway enrichment analysis

To compare the differences in biological states and pathways between cell subtypes, GO, KEGG, Gene set Enrichment analysis (GSEA) or Gene set variation analysis (GSVA) were performed. GO analysis was used to provide insights into molecular function, biological process, and cellular component of the gene sets in specific cell subpopulations. KEGG analysis was performed to identify the gene sets-enriched biological pathways. By utilizing predefined gene sets, such as metabolic pathways and signaling pathways, GSEA analysis was performed to identify whether the differential gene expression between two different cell subtypes was significantly enriched in these gene sets. The normalized enrichment score (NES) represented the enrichment degree of gene sets. GSVA analysis was performed to reveal functional differences between different cell lineages.

Trajectory analysis

To map differentiation within the gastric carcinogenesis microenvironment, pseudotime analysis was performed with Monocle 2 or 3 to determine the dynamic differentiation trajectories of cell subpopulations and gene expression changes at different gastric lesions. Utilizing the Monocle R package, genes exhibiting significant expression variation across cells were firstly identified, and their expression profiles were used for dimensionality reduction. Subsequently, a minimum spanning tree (MST) was constructed, representing the differentiation trajectories. Compared to Monocle 2, the Monocle 3 software package incorporates novel data procession methods, such as Destiny, and trajectory analysis algorithm DDRTree.

SCENIC analysis

Gene regulatory networks were constructed by SCENIC analysis. The analysis of transcriptional factor binding motifs was performed using RcisTarget R package. Grnboost was used to infer the co-expression network. Finally, the co-expressed modules between transcriptional factors and potential target genes, as well as the regulon activity score (RAS) of single cell. The regulon specificity score (RSS) inferred the specific correspondence of regulon to each cell. Connection specificity index (CSI) represented the association between different regulons.

Intercellular interaction analysis

To analyze cell-cell interaction, we applied the CellChatDB R package to identify significant ligand-receptor pairs between different cell types, including all cell subtypes, fibroblasts-tumor cells, EC-tumor cells, fibroblasts-myeloid cells, EC-myeloid cells, within all gastric tissue samples. We also compared the cell-cell interactions between H. pylori -positive and H. pylori -negative samples.

GEO and TCGA database analysis

To validate gene expression levels in bulk RNA-seq, TCGA-STAD (The Cancer Genome Atlas-Stomach carcinoma) dataset from the National Cancer Institute, and a series of RNA-seq datasets of GC and paired adjacent normal tissues (GSE122401, GSE65801 and GSE13195) from GEO database. Firstly, RNA-seq data was retrieved from public database and performed quality control and data cleaning to ensure the reliability of the results. The ggplot2 R package was then used to generate box plots illustrating the gene expression patterns in cancer and adjacent tissues.

Animal studies

Six to eight-week male specific-pathogen-free (SPF) INS-GAS mice were purchased from Jackson Lab. All animal experiments were approved by the Ethics Committee of First Affiliated Hospital of Nanchang University. After 1 week of acclimation, mice were infected with the mouse-adapted wild-type H. pylori strain (PMSS1) (2×10⁹ CFU/mouse) every other day for five times. The uninfected control mice were challenged with Brucella Broth. All mice were euthanized at 16 weeks post infection.

Animal sample collection and histological analysis

Following euthanasia, the proximal duodenum was removed and the stomach incision was made along the greater curvature. Linear gastric strips from the lesser curvature were fixed overnight in formalin. Then, the tissue sections were dehydrated, embedded and stained with H&E. Gastric histopathological scoring of gastric corpus for inflammation, gastric defect, atrophy and intestinal metaplasia (IM) was conducted by pathologists blined to sample identity. The pathological severity was graded on an ascending scale from 0 to 4. Representative histopathological images were captured under an upright fluorescent microscope (Nikon).

Confirmation of H. pylori colonization in INS-GAS mice

Gastric tissue sections stored in -80℃ freezer were cultured on selective media under microaerophilic conditions at 37℃. The morphology and motility of the bacteria were observed in microscope. Additionally, silver staining of gastric tissues was performed to further confirm the colonization of H. pylori.

Immunohistochemistry (IHC) staining

IHC analysis was performed to assess protein expression in human gastric tissues, as described previously(68). All tissues were collected from the patients with 30 of non-atrophic gastritis, 30 of IM, and 30 of GC at The First Affiliated Hospital of Nanchang University. Each lesion stage was divided into half of the positive and negative specimens of H. pylori. Briefly, tissue sections were deparaffinized in a 70℃ oven for 1.5h, followed by rehydration using a graded ethanol series. Antigen retrieval was performed by boiling the sections in citrate buffer in a microwave for 15 min. Endogenous peroxidase activity was quenched by the incubation in 3% H₂O₂ for 10min. The sections were then incubated with the primary antibodies overnight at 4℃. For a full list of primary and secondary antibodies used for immunohistochemistry, see Table S9. Subsequently, tissue sections were incubated with the secondary antibody and visualized using a DAB staining kit according to the manufactures’ instructions (Zhongshan Biotech, Beijing, China). All images were acquired under a microscope (Nikon Eclipse).

Immunofluorescence staining

The tissue sections underwent deparaffinization and rehydration steps, followed by antigen retrieval in citrate buffer (pH 6.0). Then tissue sections were permeabilized with 0.3% Triton X-100 in PBS for 15 min, followed by blocking with 3% BSA in PBS for 1 hour. The primary antibodies were then applied overnight at 4℃, see Table S9. For immunofluorescence staining, the tissue slides were washed in PBS and incubated with Alexa-flour secondary antibodies (Invitrogen) at room temperature for 60 min, then mounted with an appropriate medium containing 4',6-diamidino-2-phenylindole (DAPI) nuclear counterstain. The mounted sections were then imaged using a confocal microscope (LEICA Stellaris 5).

qRT-PCR analysis

Gastric biopsy tissues containing 30 of GS, 30 of IM and 24 of GC were derived from the remaining specimens from the previous study(69). Each gastric lesion stage was divided into infected and uninfected groups based on H. pylori infection status. The qRT-PCR analysis was performed to detect the mRNA levels of target genes, as described previously(68). Briefly, total RNA was extracted from the tissue samples using Trizol reagent (TIANGEN Biotech, Beijing, China). Then RNA was converted into cDNA using a reverse transcription kit according to the manufacturer’s instructions. The qPCR assay was performed with a QuantStudio 5 Real-time PCR system (Life Technologies). The primers for qRT-PCR analysis used in this study were listed in Table S10.

Statistical analysis

All statistical analyses and graph were performed using SPSS software (V 21.0) and graphpad prism 9. The Wilcoxon rank-sum test was used to compare continuous variables between two independent samples. The t-test was used to evaluate association with parametric variables between two independent samples. The results were considered statistically significant at p＜0.05 (*, p＜0.05; **, p＜0.01; ***, p＜0.001).

Ethics approval and consent to participate

This work was approved by the Ethics Committee of the First Affiliated Hospital of Nanchang University under protocol number of (2023) CDYFYYLK (01-009). All tissue samples were obtained under informed consent.

Consent for publication

Not applicable

Availability of data and materials

The raw single-cell sequencing data of 18 human gastric samples from this study was deposited in the BIG Sub system under BioProject accession number PRJCA018947.

Competing interests

The authors declare no competing interests.

Funding

This work was supported by the National Natural Science Foundation of China (82260119, 82170580 and 82260118), Academic and Technical Leader of major disciplines in Jiangxi Province (20225BCJ23021) and the Natural Science Foundation of Jiangxi Province (20212BAB216016, 20224ACB216004).

Author contribution

Conceptualization, Y. Zhu, N. Lu, and N. Li; Methodology, Y. Zhu, N. Lu, N. Li, X. Xu, and C. He; Tissue samples collection, N. Li, P. Zheng, Y. Zhou, X. Yang, Y. Ouyang, Y. Hu, Y. Cao, Z. Li, and Y. Zhu; Software and bioinformatic analysis, N. Li, S. Chen, J. Liu, P. Zheng and C. Xie; Validation, N. Li, X. Xu, P. Zheng, Y. Lei, and X. Fei; Writing-original Draft: N. Li, X. B; Writing-Review& Editing: N. Li, J. Liu, J. Guo, William KK, Z. Ge, N. Lu, and Y. Zhu; Visualization: N. Li and J. Liu; Funding Acquisition: N. Li, J. Liu, Y. Zhu and N. Lu; Supervision: Y. Zhu, N. Lu and J. Liu.

Acknowledgements

We thank all the patients for their generous donation of tissue samples in this study. We also thank Dr. Yongbing Ba and Wenyang Ding from OE biotech company (Shanghai, China) for the support of bioinformatic analysis.

Arnold M, Ferlay J, van Berge Henegouwen MI, and Soerjomataram I. Global burden of oesophageal and gastric cancer by histology and subsite in 2018. Gut. 2020;69(9):1564–71.
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, and Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424.
Crowe SE. Helicobacter pylori Infection. N Engl J Med. 2019;380(12):1158–65.
Correa P, and Piazuelo MB. Helicobacter pylori Infection and Gastric Adenocarcinoma. US Gastroenterol Hepatol Rev. 2011;7(1):59–64.
Sun D, Lei L, Xia C, Li H, Cao M, He S, et al. Sociodemographic disparities in gastric cancer and the gastric precancerous cascade: A population-based study. Lancet Reg Health West Pac. 2022;23:100437.
Yan L, Chen Y, Chen F, Tao T, Hu Z, Wang J, et al. Effect of Helicobacter pylori Eradication on Gastric Cancer Prevention: Updated Report From a Randomized Controlled Trial With 26.5 Years of Follow-up. Gastroenterology. 2022.
Cancer Genome Atlas Research N. Comprehensive molecular characterization of gastric adenocarcinoma. Nature. 2014;513(7517):202–9.
Mun DG, Bhin J, Kim S, Kim H, Jung JH, Jung Y, et al. Proteogenomic Characterization of Human Early-Onset Gastric Cancer. Cancer Cell. 2019;35(1):111–24 e10.
Kumar V, Ramnarayanan K, Sundar R, Padmanabhan N, Srivastava S, Koiwa M, et al. Single-Cell Atlas of Lineage States, Tumor Microenvironment, and Subtype-Specific Expression Programs in Gastric Cancer. Cancer Discov. 2022;12(3):670–91.
Zhang M, Hu S, Min M, Ni Y, Lu Z, Sun X, et al. Dissecting transcriptional heterogeneity in primary gastric adenocarcinoma by single cell RNA sequencing. Gut. 2020.
Sathe A, Grimes SM, Lau BT, Chen J, Suarez C, Huang RJ, et al. Single-Cell Genomic Characterization Reveals the Cellular Reprogramming of the Gastric Tumor Microenvironment. Clin Cancer Res. 2020;26(11):2640–53.
Liabeuf D, Oshima M, Stange DE, and Sigal M. Stem Cells, Helicobacter pylori, and Mutational Landscape: Utility of Preclinical Models to Understand Carcinogenesis and to Direct Management of Gastric Cancer. Gastroenterology. 2022;162(4):1067–87.
Li N, Feng Y, Hu Y, He C, Xie C, Ouyang Y, et al. Helicobacter pylori CagA promotes epithelial mesenchymal transition in gastric carcinogenesis via triggering oncogenic YAP pathway. J Exp Clin Cancer Res. 2018;37(1):280.
Xie C, Li N, Wang H, He C, Hu Y, Peng C, et al. Inhibition of autophagy aggravates DNA damage response and gastric tumorigenesis via Rad51 ubiquitination in response to H. pylori infection. Gut Microbes. 2020;11(6):1567–89.
Sun Y, Wu L, Zhong Y, Zhou K, Hou Y, Wang Z, et al. Single-cell landscape of the ecosystem in early-relapse hepatocellular carcinoma. Cell. 2021;184(2):404–21 e16.
Longman RJ, Douthwaite J, Sylvester PA, Poulsom R, Corfield AP, Thomas MG, et al. Coordinated localisation of mucins and trefoil peptides in the ulcer associated cell lineage and the gastrointestinal mucosa. Gut. 2000;47(6):792–800.
Bockerstett KA, Lewis SA, Noto CN, Ford EL, Saenz JB, Jackson NM, et al. Single-Cell Transcriptional Analyses Identify Lineage-Specific Epithelial Responses to Inflammation and Metaplastic Development in the Gastric Corpus. Gastroenterology. 2020.
Bockerstett KA, Lewis SA, Wolf KJ, Noto CN, Jackson NM, Ford EL, et al. Single-cell transcriptional analyses of spasmolytic polypeptide-expressing metaplasia arising from acute drug injury and chronic inflammation in the stomach. Gut. 2020;69(6):1027–38.
Gehart H, van Es JH, Hamer K, Beumer J, Kretzschmar K, Dekkers JF, et al. Identification of Enteroendocrine Regulators by Real-Time Single-Cell Differentiation Mapping. Cell. 2019;176(5):1158-73 e16.
Chen HY, Hu Y, Lu NH, and Zhu Y. Caudal type homeoboxes as a driving force in Helicobacter pylori infection-induced gastric intestinal metaplasia. Gut Microbes. 2020;12(1):1–12.
Miao ZF, Sun JX, Adkins-Threats M, Pang MJ, Zhao JH, Wang X, et al. DDIT4 Licenses Only Healthy Cells to Proliferate During Injury-induced Metaplasia. Gastroenterology. 2021;160(1):260–71 e10.
Chen L, Vasoya RP, Toke NH, Parthasarathy A, Luo S, Chiles E, et al. HNF4 Regulates Fatty Acid Oxidation and Is Required for Renewal of Intestinal Stem Cells in Mice. Gastroenterology. 2020;158(4):985–99 e9.
Aguilar C, Pauzuolis M, Pompaiah M, Vafadarnejad E, Arampatzi P, Fischer M, et al. Helicobacter pylori shows tropism to gastric differentiated pit cells dependent on urea chemotaxis. Nat Commun. 2022;13(1):5878.
Sharma D, Malik A, Guy CS, Karki R, Vogel P, and Kanneganti TD. Pyrin Inflammasome Regulates Tight Junction Integrity to Restrict Colitis and Tumorigenesis. Gastroenterology. 2018;154(4):948–64 e8.
Oh HK, Tan AL, Das K, Ooi CH, Deng NT, Tan IB, et al. Genomic loss of miR-486 regulates tumor progression and the OLFM4 antiapoptotic factor in gastric cancer. Clin Cancer Res. 2011;17(9):2657–67.
Rogers AB, Taylor NS, Whary MT, Stefanich ED, Wang TC, and Fox JG. Helicobacter pylori but not high salt induces gastric intraepithelial neoplasia in B6129 mice. Cancer Res. 2005;65(23):10709–15.
Palikuqi B, Nguyen DT, Li G, Schreiner R, Pellegata AF, Liu Y, et al. Adaptable haemodynamic endothelial cells for organogenesis and tumorigenesis. Nature. 2020;585(7825):426–32.
Hayakawa Y, Nakagawa H, Rustgi AK, Que J, and Wang TC. Stem cells and origins of cancer in the upper gastrointestinal tract. Cell Stem Cell. 2021;28(8):1343–61.
Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, et al. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088.
Kobayashi H, Enomoto A, Woods SL, Burt AD, Takahashi M, and Worthley DL. Cancer-associated fibroblasts in gastrointestinal cancer. Nat Rev Gastroenterol Hepatol. 2019;16(5):282–95.
Buechler MB, Pradhan RN, Krishnamurty AT, Cox C, Calviello AK, Wang AW, et al. Cross-tissue organization of the fibroblast lineage. Nature. 2021;593(7860):575–9.
Lovatt D, Tamburino A, Krasowska-Zoladek A, Sanoja R, Li L, Peterson V, et al. scRNA-seq generates a molecular map of emerging cell subtypes after sciatic nerve injury in rats. Commun Biol. 2022;5(1):1105.
Zhang M, Yang H, Wan L, Wang Z, Wang H, Ge C, et al. Single-cell transcriptomic architecture and intercellular crosstalk of human intrahepatic cholangiocarcinoma. J Hepatol. 2020;73(5):1118–30.
Liu J, Xiao Q, Xiao J, Niu C, Li Y, Zhang X, et al. Wnt/beta-catenin signalling: function, biological mechanisms, and therapeutic opportunities. Signal Transduct Target Ther. 2022;7(1):3.
Liu T, Liu C, Yan M, Zhang L, Zhang J, Xiao M, et al. Single cell profiling of primary and paired metastatic lymph node tumors in breast cancer patients. Nat Commun. 2022;13(1):6823.
Mao X, Xu J, Wang W, Liang C, Hua J, Liu J, et al. Crosstalk between cancer-associated fibroblasts and immune cells in the tumor microenvironment: new findings and future perspectives. Mol Cancer. 2021;20(1):131.
Zilionis R, Engblom C, Pfirschke C, Savova V, Zemmour D, Saatcioglu HD, et al. Single-Cell Transcriptomics of Human and Mouse Lung Cancers Reveals Conserved Myeloid Populations across Individuals and Species. Immunity. 2019;50(5):1317-34 e10.
Barata JT, Durum SK, and Seddon B. Flip the coin: IL-7 and IL-7R in health and disease. Nat Immunol. 2019;20(12):1584–93.
Xu W, Zhao X, Wang X, Feng H, Gou M, Jin W, et al. The Transcription Factor Tox2 Drives T Follicular Helper Cell Development via Regulating Chromatin Accessibility. Immunity. 2019;51(5):826–39 e5.
Guo X, Zhang Y, Zheng L, Zheng C, Song J, Zhang Q, et al. Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing. Nat Med. 2018;24(7):978–85.
Zheng C, Zheng L, Yoo JK, Guo H, Zhang Y, Guo X, et al. Landscape of Infiltrating T Cells in Liver Cancer Revealed by Single-Cell Sequencing. Cell. 2017;169(7):1342-56 e16.
Szabo PA, Levitin HM, Miron M, Snyder ME, Senda T, Yuan J, et al. Single-cell transcriptomics of human T cells reveals tissue and activation signatures in health and disease. Nat Commun. 2019;10(1):4706.
Sasson SC, Slevin SM, Cheung VTF, Nassiri I, Olsson-Brown A, Fryer E, et al. Interferon-Gamma-Producing CD8(+) Tissue Resident Memory T Cells Are a Targetable Hallmark of Immune Checkpoint Inhibitor-Colitis. Gastroenterology. 2021;161(4):1229–44 e9.
Salome B, Sfakianos JP, Ranti D, Daza J, Bieber C, Charap A, et al. NKG2A and HLA-E define an alternative immune checkpoint axis in bladder cancer. Cancer Cell. 2022;40(9):1027–43 e9.
Mills KHG. IL-17 and IL-17-producing cells in protection versus pathology. Nat Rev Immunol. 2023;23(1):38–54.
Liu Y, Zhang Q, Xing B, Luo N, Gao R, Yu K, et al. Immune phenotypic linkage between colorectal cancer and liver metastasis. Cancer Cell. 2022;40(4):424–37 e5.
Miheecheva N, Postovalova E, Lyu Y, Ramachandran A, Bagaev A, Svekolkin V, et al. Multiregional single-cell proteogenomic analysis of ccRCC reveals cytokine drivers of intratumor spatial heterogeneity. Cell Rep. 2022;40(7):111180.
Togashi Y, Shitara K, and Nishikawa H. Regulatory T cells in cancer immunosuppression - implications for anticancer therapy. Nat Rev Clin Oncol. 2019;16(6):356–71.
Plitas G, Konopacki C, Wu K, Bos PD, Morrow M, Putintseva EV, et al. Regulatory T Cells Exhibit Distinct Features in Human Breast Cancer. Immunity. 2016;45(5):1122–34.
Wang R, Song S, Qin J, Yoshimura K, Peng F, Chu Y, et al. Evolution of immune and stromal cell states and ecotypes during gastric adenocarcinoma progression. Cancer Cell. 2023;41(8):1407–26 e9.
Kalliolias GD, and Ivashkiv LB. TNF biology, pathogenic mechanisms and emerging therapeutic strategies. Nat Rev Rheumatol. 2016;12(1):49–62.
Morgan E, Arnold M, Camargo MC, Gini A, Kunzmann AT, Matsuda T, et al. The current and future incidence and mortality of gastric cancer in 185 countries, 2020-40: A population-based modelling study. EClinicalMedicine. 2022;47:101404.
Wang R, Dang M, Harada K, Han G, Wang F, Pool Pizzi M, et al. Single-cell dissection of intratumoral heterogeneity and lineage diversity in metastatic gastric adenocarcinoma. Nat Med. 2021;27(1):141–51.
Huang XZ, Pang MJ, Li JY, Chen HY, Sun JX, Song YX, et al. Single-cell sequencing of ascites fluid illustrates heterogeneity and therapy-induced evolution during gastric cancer peritoneal metastasis. Nat Commun. 2023;14(1):822.
Zhang P, Yang M, Zhang Y, Xiao S, Lai X, Tan A, et al. Dissecting the Single-Cell Transcriptome Network Underlying Gastric Premalignant Lesions and Early Gastric Cancer. Cell Rep. 2019;27(6):1934–47 e5.
Dissecting the Single-Cell Transcriptome Network Underlying Gastric Premalignant Lesions and Early Gastric Cancer - PubMed.
Chen L, Toke NH, Luo S, Vasoya RP, Fullem RL, Parthasarathy A, et al. A reinforcing HNF4-SMAD4 feed-forward module stabilizes enterocyte identity. Nat Genet. 2019;51(5):777–85.
Zhang P, Yang M, Zhang Y, Xiao S, Lai X, Tan A, et al. Dissecting the Single-Cell Transcriptome Network Underlying Gastric Premalignant Lesions and Early Gastric Cancer. Cell Rep. 2020;30(12):4317.
Lynch MD, and Watt FM. Fibroblast heterogeneity: implications for human disease. J Clin Invest. 2018;128(1):26–35.
Chen Y, McAndrews KM, and Kalluri R. Clinical and therapeutic relevance of cancer-associated fibroblasts. Nat Rev Clin Oncol. 2021;18(12):792–804.
Elyada E, Bolisetty M, Laise P, Flynn WF, Courtois ET, Burkhart RA, et al. Cross-Species Single-Cell Analysis of Pancreatic Ductal Adenocarcinoma Reveals Antigen-Presenting Cancer-Associated Fibroblasts. Cancer Discov. 2019;9(8):1102–23.
Huang H, Wang Z, Zhang Y, Pradhan RN, Ganguly D, Chandra R, et al. Mesothelial cell-derived antigen-presenting cancer-associated fibroblasts induce expansion of regulatory T cells in pancreatic cancer. Cancer Cell. 2022;40(6):656–73 e7.
Chaturvedi R, Asim M, Hoge S, Lewis ND, Singh K, Barry DP, et al. Polyamines Impair Immunity to Helicobacter pylori by Inhibiting L-Arginine Uptake Required for Nitric Oxide Production. Gastroenterology. 2010;139(5):1686–98, 98 e1-6.
Zhang Q, He Y, Luo N, Patel SJ, Han Y, Gao R, et al. Landscape and Dynamics of Single Immune Cells in Hepatocellular Carcinoma. Cell. 2019;179(4):829–45 e20.
Moncada R, Barkley D, Wagner F, Chiodin M, Devlin JC, Baron M, et al. Integrating microarray-based spatial transcriptomics and single-cell RNA-seq reveals tissue architecture in pancreatic ductal adenocarcinomas. Nat Biotechnol. 2020;38(3):333–42.
Ozga AJ, Chow MT, and Luster AD. Chemokines and the immune response to cancer. Immunity. 2021;54(5):859–74.
Lee SH, Contreras Panta EW, Gibbs D, Won Y, Min J, Zhang C, et al. Apposition of Fibroblasts With Metaplastic Gastric Cells Promotes Dysplastic Transition. Gastroenterology. 2023;165(2):374–90.
Li N, Xu X, Zhan Y, Fei X, Ouyang Y, Zheng P, et al. YAP and beta-catenin cooperate to drive H. pylori-induced gastric tumorigenesis. Gut Microbes. 2023;15(1):2192501.
He C, Peng C, Shu X, Wang H, Zhu Z, Ouyang Y, et al. Convergent dysbiosis of gastric mucosa and fluid microbiome during stomach carcinogenesis. Gastric Cancer. 2022;25(5):837–49.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Single-cell transcriptomic profiling uncovers cellular complexity and microenvironment in gastric tumorigenesis associated with Helicobacter pylori

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Materials and Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1