Single-cell and spatial transcriptomic analyses reveal the cellular origins and drivers of liver metastasis from pancreatic cancer

Single-cell and spatial transcriptomic analyses reveal the cellular origins and drivers of liver metastasis from pancreatic cancer | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Single-cell and spatial transcriptomic analyses reveal the cellular origins and drivers of liver metastasis from pancreatic cancer Jing Cui, Yao Guo, Taoyu Chen, Yan Ling, Xueyi Liang, Jingyuan Zhao, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4128922/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Pancreatic cancer is a deadliest malignancy. The most common event of tumor progression in pancreatic cancer is liver metastasis, and current treatments have not produced the desired results. However, the cellular origins and drivers of liver metastasis have yet to be defined. In this study, we identified nine cell clusters in pancreatic cancer via analyzing single-cell transcriptomic profiles. Malignant epithelial cells in liver metastatic group might come from primary tumor subclones with 27 common gene copy number variation. Cell-cell communication demonstrated that malignant epithelial cells were specifically regulated by IL-1, Periostin, and IFN-II signaling pathways and fibroblasts regulated malignant epithelial cells through the Periostin pathway (POSTN-ITGAV/ITGB5). Moreover, we found that 34 differentially expressed genes were highly enriched in PI3K signaling pathway. Next, we constructed a LM index using these 34 genes, and found that pancreatic cancer patients with high-index exhibited a poor prognosis. The immuno-fluorescence staining assay showed that the spatial distance between POSTN and ITGAV was close, and ITGAV protein was expressed in tissue sections of pancreatic cancer patients. Finally, we found that ITGAV was upregulated in PANC-1 cells, and knockdown of ITGAV inhibited the invasion and migration of PANC-1 cells. This study clarified possible cellular origins and driv-ers of LM pancreatic cancer based on the single-cell and spatial transcriptomic profiles, and dis-covered that the ITGAV mRNA and protein were expressed in pancreatic cancer tissue and cells, and the knockdown of ITGAV inhibited the progression of pancreatic cancer. Biological sciences/Genetics/Genomics/Transcriptomics Biological sciences/Cancer/Metastases ITGAV liver metastasis malignant epithelial cells pancreatic cancer single cell sequencing Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1 Introduction Pancreatic cancer is currently the seventh leading cause of cancer-related death worldwide, and it is also the malignant tumor with the mortality closest to the incidence, which is characterized by high recurrence rate and low survival rate 1, 2 . According to statistics, in 2022, there were 134374 new patients with pancreatic cancer and 131203 deaths from pancreatic cancer in China 3 . The overall population incidence ranked eighth in the tumor category, and the mortality ranked sixth in the tumor category. Among them, the male incidence ranked fifth and the mortality ranked third, both higher than the female, showing certain gender differences 3 . Pancreatic cancer risk factors include smoking, chronic pancreatitis, obesity, long-term diabetes, a significant family history of pancreatic cancer, and diets heavy in red and processed meats 4 . Currently, surgery is the most effective treatment for localized pancreatic cancer 5, 6 , but, less than 20% of patients are suitable for curative resection at the first diagnosis, and chemotherapy remains the cornerstone of treatment for both unresectable advanced and metastatic pancreatic cancer 7 . It has been reported that the most common metastatic site of pancreatic cancer is the liver, and liver metastasis (LM) occurs in about 70% of pancreatic cancer patients 8, 9 , and the average survival time of pancreatic cancer patients with liver metastases is about 3 to 6 months 10 . Accordingly, it is particularly important to characterize the molecular characteristics of LM of pancreatic cancer, which may help to find a new scheme for the diagnosis of early pancreatic cancer. In recent years, the application of single-cell sequencing data has revealed dynamic changes in the tumor microenvironment (TME) during malignant progression of pancreatic cancer 11 . Sang et al. have found that CD8+T cells are the immune cell type most severely affected by ripk2 deficiency in the TME of pancreatic ductal adenocarcinoma (PDAC) 12 . Zhao et al. have collected the whole transcriptome data of 1200 PDAC patients. Through retrospective meta-analysis, they divided PDAC into six molecular subtypes, each of which has its characteristic gene expression profile and shows different clinical characteristics, so as to better stratify PDAC patients and provide personalized treatment 13 . By analyzing the single-cell RNA-seq data of 24 primary PDAC and 11 normal pancreases, Peng et al. have found that the proliferation of subpopulations with high expression of cancer cells associated with malignant behavior is often accompanied by the loss of activated T cells, which predicts a poor prognosis 14 . Moreover, Ligorio et al. have detected notable changes in single-cell populations towards proliferative (PRO) and invasive epithelial-to-mesenchymal transition (EMT) phenotypes in pancreatic cancer, which are associated with signal transducer and activator of transcription 3 (STAT3) and mitogen-activated protein kinase (MAPK) signaling by using single-cell RNA sequencing technology and RNA in situ hybridization technologies 15 . Therefore, analyzing the single-cell RNA-seq data of pancreatic cancer will help determine the cell subtypes and explore the therapeutic strategies for pancreatic cancer. This study focused on LM from pancreatic cancer. By analyzing single-cell transcriptome data from PT and LM pancreatic cancer samples, we identified nine cell clusters and malignant epithelial cells. Moreover, we constructed a metastasis-related prognosis signature in pancreatic cancer, and analyzed spatial distance between POSTN and ITGAV in pancreatic cancer based on the spatial transcriptomic profiles. Finally, we explored the role of ITGAV in the progression of pancreatic cancer. Our findings are expected to provide further insights into the development of therapeutic strategies for pancreatic cancer. 2 Materials And Methods 2.1 Data collection Single cell sequencing data of pancreatic cancer were downloaded from Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/, ID: GSE154778 and GSE154778) database. GSE156405 contained 5 primary tumor (PT) and 1 liver metastatic (LM) pancreatic cancer samples, and GSE154778 included 10 PT samples and 5 LM samples (Table S1). The transcriptome data of 183 pancreatic cancer samples were retrieved from The Cancer Genome Atlas (TCGA, https://tcga-data.nci.nih.gov/tcga/) database. Among which, totally 126 samples had complete survival information (Table S2). Furthermore, spatial transcriptome data of PT samples (ST_PDAC-1 (GSM6132061)) by high throughput sequencing were extracted from GSE202740 dataset. 2.2 Single cell data quality control and cell annotation Cell Ranger (v6.1.2) was utilized to align the data to the human genome (GRCh37). Seurat v4.1.1 was used to process single-cell data, and cells with mitochondrial content higher than 20%, hemoglobin content higher than 5%, and expression genes less than 200 were filtered. Seurat was used for data normalization, cell clustering and dimension reduction. The "FindVariableFeatures" function was used to select 2000 highly variable genes from the corrected expression matrix, and then the "RunPCA" function was used for principal component analysis, retaining the top 20 principal components for further analysis. Batch effects were corrected by "RunHarmony" of R package harmony. The "FindClusters" function was used for cell clustering (resolution 0.6), and the "RunUMAP" function was used for nonlinear dimensionality reduction. Cell grouping was annotated based on cellmark2.0 database and common mark genes of cells. 2.3 Differential gene and functional enrichment analyses The “Findmarkers” function from the Seurat package was utilized to performed differential gene analysis. The differentially expressed genes (DEGs) among between two groups were screened by p.adj < 0.05. The R package "ClusterProfiler" was used for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome (KEGG) enrichment analysis. The significantly enriched pathways were screened using p < 0.05. 2.4 Gene set variation analysis (GSVA) The Hallmark gene set was downloaded for GSVA from the MSigDB(v2023.1) (https://www.gsea-msigdb.org) database to investigate the difference of the biological function between two groups using R package. The difference analysis among groups were conducted using the "limma" package. The differentially enriched pathways between two groups were identified based on |t| > 2 and p < 0.05. 2.5 Copy number variation (CNV) analysis Based on single-cell gene expression and chromosome sequencing data, the inferCNV (v1.14.0) package in R was used to distinguish malignant epithelial cells from non-malignant epithelial cells. The inferCNV analysis was performed with the following settings: cutoff=0.1, cluster-by-groups=TRUE, tumor subcluster-partition-method = "random-trees", and hidden markov model (HMM) = TRUE. To reduce false positive calls in CNV inference, the default Bayesian latent mixture model was applied to determine the posterior probability of changes in each cell, with a default threshold of 0.5. Furthermore, all gene CNV scores of ECs and reference cells (T cells) were hierarchically clustered using the k-means algorithm. The malignant epithelial cells were identified according to the CNV score. If the score was greater than the 95th percentile of the reference cell, it is malignant epithelial cells. The others were non-malignant epithelial cells. To illustrate tumor clonality and evolution, the "sub cluster" pattern was further applied to divide malignant cells into eight clusters, and HMM generated different CNV patterns. Each CNV was annotated as a gain or loss at the p or q arm level according to UCSC chromosome band information. Subclones containing the same arm level CNVs were collapsed to construct an evolutionary tree. The phylogenetic tree was drawn using Uphyloplot2 software to represent the subclonal CNV structure. 2.6 Inference of developmental trajectory The Monocle (v2.28.0) was used to construct pesudotime trajectories based on gene expression profiles of malignant epithelial cells. After dimensionality reduction and cell sorting, all malignant epithelial cells were projected and sorted into trajectories with different branches, and cells within the same branch were considered to have the same cell state. Branch expression analysis modeling (BEAM) was further performed to identify genes with branch dependent expression patterns. These branch dependent genes can help us to explore the mechanism of cell fate determination. 2.7 Cell communication analysis Cellchat (v1.5.0) package was used to predict and visualize biologically meaningful intercellular communication. For each individual dataset, first extracted the expression matrix and metadata from the Seurat object, and then used the "createCellChat" function to generate a CellChat object. After calculating the highly variable genes and pathways, the "computeCommunProb" function was used to infer the intercellular communication probability. The results were shown using a series of visualization functions provided by CellChat, such as "netVisual-bubble" presented the dot diagram of signaling pathways emitted from cells. The functions "compareInteractions", "netVisual-diffinteraction", "netVisual-heatmap" of the CellChat package were used to compare the relative number or interaction strength of various cell subsets between two groups. 2.8 LM-index and overall survival analysis Based on the risk genes, the gelnet (v1.2.1) package in R was applied to construct a LM-index by one class logistic regression (OCLR) algorithm to represent the risk of LM pancreatic cancer samples. The patients were divided into high and low LM-index groups using the R package "survminer". Kaplan Meier (K-M) survival analysis was used to evaluate the overall survival (OS) of patients with high and low LM-index, and two-sided log rank test was used for comparison. 2.9 Spatial data analysis Raw data was quality controlled using fastp, and then data alignment and count were performed using SpaceRanger (v2.1.0). Seurat package (v4.4.0) was used for data filtering (spot with more than 200 retained genes, less than 20% mitochondria, and less than 5% hemoglobin), spot normalization and regression. Finally, Seurat's FindTransferAnchors and transferdata functions were employed to map single-cell data and achieve annotation of spatial data. 2.10 Immunofluorescence staining Pancreatic cancer tissue microarray samples were purchased from bioaitech (Xi’an,China). The use of the human tissue microarrays was approved by the Ethics Committee of Union Hospital of Tongji Medical College of Huazhong University of Science and Technology. The in-formation of patients was shown in Table S3. ITGAV (Proteintech, #27096-1-ap) and postn (Proteintech, #66491-1-Ig) were used as primary antibodies. Immunofluorescence staining was performed according to the instructions of alphatsa multiplex IHC Kit (AXT37100031, Alphaxbio). In brief, the tissue chip was dewaxed and hydrated through a series of xylene and alcohol washes, and then antigen repair and sealing were performed. The sections were blocked and incubated with primary antibody and secondary antibody, and fluorescent staining was performed. Finally, the nuclei were counterstained with DAPI for 5 min and enclosed in Mounting Medium. ZEN (v3.1) software was used for film reading. 2.11 Cell collection and culture Human pancreatic cancer cell line PANC-1 and human normal pancreatic ductal epithelial cell line HPDE6-C7 were purchased from BeNa culture collection (Beijing, China). All cells were cultured in RPMI-1640 complete medium (Gibco) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. The cells were placed in a humidified incubator maintained at 37 °C with 5% CO 2 . 2.12 qRT-PCR Total of RNA from cells was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA), and the quality and concentration of RNA were evaluated using UV spectrophotometer, and then reverse transcription was performed using Transcriptor First Strand cDNA Synthesis Kit (GenStar, Beijing, China). Furthermore, qPCR detection was performed using the LightCycler 480 fluorescence quantification system (Roche, Basel, Switzerland). The reference gene was GADPH, and the primer sequences were listed in Table S4. The mRNA expression levels were calculated using the 2-ΔΔCT method (three repeats). 2.13 Cell transfection The small interfering RNA (siRNA) was used to knock down ITGVA in PANC-1 cells. The ITGAV siRNA sequence was presented in Table S5. The cells were seeded onto 6-well plates and transfected once they reached approximately 50% confluence. The transfection involved using a negative control (NC) and si-ITGAV to transfect the cells, respectively. All transfections were carried out with Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). The knockdown efficiency of ITGAV was detected by qRT-PCR. 2.14 Cell scratch assay A total of 7×10 5 cells were seeded in each well of a 6-well and cultured for 24 hours. A line was drawn in the center of the well using a 10 μL pipette tip. After two washes with PBS, the cells were cultured in a 37 °C incubator for 24 hours. Subsequently, the wounds were photographed using a microscope at various time intervals. The distances of the wounds were measured using Fiji (ImageJ). 2.15 Invasion and migration assays The migration and invasion abilities of cells were analyzed using a polycarbonate membrane with an 8 μm pore size in a 24-well Transwell chamber (Coring, NY, USA). The upper chamber added 1×10 4 cells in serum-free medium containing 0.1% BSA. Moreover, the lower chamber was supplemented with medium containing 0.1% BSA and EGF (50 ng/ml, MCE, NJ, USA). After incubation for 24 h, the cells in the upper chamber were completely transferred to the lower membrane. The polycarbonate membrane was fixed and stained with Giemsa solution (Solarbio, Beijing, China), and photographed with a microscope. 3 Results 3.1 Characterization of discrete cellular compositions in pancreatic cancer via meticulous single-cell analysis Firstly, we extracted totally 35682 cells from PT and LM pancreatic cancer samples, including 17961 PT cells and 17721 LM pancreatic cancer cells. Subsequently, these 35682 cells were clustered into 9 clusters by cell clustering: Epithelial cells (KRT19), T cells (CD3D), macrophage (SPP1), fibroblast (COL3A1), monocyte (S100A9), natural killer (NK: NKG7), B cells (CD79A), endothelial (COL4A1/CLDN5), dendritic cell (DC: STMN1/JCHAIN) (Figure 1A-1F). The proportion of epithelial cells and T cells were higher in PT and LM pancreatic cancer samples. Compared to LM samples, the proportion of fibroblast and macrophage were increased in PT samples (Figure 1 G). To elucidate the involvement of cell clusters in the metastatic process of pancreatic cancer, we conducted a GSVA enrichment analysis of cell clusters in the PT and LM samples. In the PT samples, the glycolysis and metabolic related pathways, such as Cholesterol homeostasis, Bile acid metabolism were significantly activated in the epithelial cell cluster, and epithelial-mesenchymal transition (EMT) and angiogenesis pathways were remarkably activated in the fibroblast cluster (Figure 1H). Compared to PT samples, Angiogenesis, Hedgehog signaling, E2F targets pathways were significantly activated in LM samples in the epithelial cell cluster, and oxidative phosphorylation and glycolysis pathways were observably activated in LM samples in the fibroblast cluster (Figure 1I). 3.2 Identification of malignant epithelial cells in pancreatic cancer To determine the clonal structure and cellular origin of malignant epithelial cell, we calculated the CNV and clonality of epithelial cells in PT and LM samples via inferCNV algorithm. Compared to the reference cells, a total of 6396 cells with higher CNV score were considered malignant epithelial cells among 6903 epithelial cells of PT samples (Figure 2A). Meanwhile, among the 8772 epithelial cells of LM samples, totally 8637 cells were considered malignant epithelial cells (Figure 2B). The proportion of malignant epithelial cells were higher in LM samples (Figure 2C, LM vs. PT). Furthermore, we utilized phylogenetic trees to illustrate the clonality of tumors and the progression of malignant cells from PT to LM pancreatic cancer samples. In all malignant epithelial cells of PT and LM samples, gain of 1q and loss of 11q were observed (Figure 2D, 2E). Totally 700 CNV changed genes were shared by 8 subclonal cell populations of PT samples (Figure 2F, Table S6), and 225 CNV changed genes were shared by all subclonal cell populations of LM samples (Figure 2G, Table S7). Moreover, we found that when pancreatic cancer tumors metastasized to the liver, copy number changes in 27 genes were retained, while LM samples also produced additional changes in 198 genes (Figure 2H). These findings indicated that malignant epithelial cells in LM might come from PT subclones with 27 common gene CNV, and malignant cells also generated more new CNV subclones in the process of transferring to liver tissue. 3.3 The results of intercellular interaction Next, we used Cellchat to analyze the cell-cell communication among malignant epithelial cells, non-malignant epithelial cells and other cells in PT samples, and the results showed that the interaction between other cells and malignant epithelial cells increased (Figure 3A-3D, Table S8-9). Compared to non-malignant epithelial cells, malignant epithelial cells were specifically regulated by IL-1, Periostin, and IFN-II signaling pathways (Figure 3E). Monocytes and macrophages regulated malignant epithelial cells through the IL-1 pathway (IL-1B-IL1R2), fibroblasts regulated malignant epithelial cells through the Periostin pathway (POSTN-ITGAV/ITGB5), and NK cells regulated malignant epithelial cells through the IFN-II pathway (IFNG-IFNGR1/IFNGR2) (Figure 3F-3I). 3.4 Construction of metastatic risk model in pancreatic cancer GO enrichment analysis showed that the DEGs between malignant and non- malignant epithelial cells were highly enriched in cadherin binding, actin binding, cell substrate junction, focal adhesion, and actin filament organization (Figure S1A). KEGG enrichment analysis showed that these DEGs were highly enriched in Tight junction, PI3K-Akt signaling pathway and Focal adhension (Figure S1B). After combined with ITGAV, POSTN can trigger changes in the intracellular PI3K signaling pathway, which leads to cancer cell proliferation and invasion 16 . Thus, we defined 34 genes enriched in PI3K pathway by malignant cells (Table S10) as risk genes and calculated liver metastatic (LM) index. According to the LM-index, the pancreatic cancer patients were divided into high and low LM-index groups. The pancreatic cancer patients with high-index exhibited a poor prognosis (Figure S1C, p =0.046). SCENIC analysis showed that ELF3, EHF, YY1, KLF2 and CREB3L1 were enriched in malignant cells (Figure S1D). 3.5 Spatial distance between POSTN and ITGAV in pancreatic cancer We mapped the single-cell data annotation results to the pancreatic cancer spatial sequencing samples and found that the samples contained fibroblasts, epithelial cells, endothelial cells and macrophages, and fibroblasts were adjacent to epithelial cells (Figure 4A). The expressions of POSTN and ITGAV in tissue sections were shown in Figure 4B. POSTN was highly expressed in fibroblast enriched areas, and ITGAV was highly expressed in epithelial cell aggregation areas. Subsequently, we used the cancer tissue sections of 6 pancreatic cancer patients for fluorescence staining analysis, and the results showed that the spatial distance between POSTN and ITGAV was close (Figure 4C). 3.6 Knockdown of ITGAV inhibited the invasion and migration of pancreatic cancer cells We found that ITGAV was expressed in tissue sections of six pancreatic cancer patients (Figure 5A), and the expression of ITGAV was significantly upregulated in PANC-1 compared with human normal pancreatic ductal epithelial cell line HPDE6-C7 (Figure 5B). In addition, to explore the effect of ITGAV in the progression of pancreatic cancer, we constructed ITGAV knockdown (si-ITGAV-1, siI-TGAV-2, si-ITGAV-3) PANC-1 cells, and discovered that ITGAV expression was significantly reduced in si-ITGAV-1, si-ITGAV-2, si-ITGAV-3 groups (Figure 6A). Moreover, we analyzed the ITGAV knockdown’s impacts on invasion and migration of PANC-1 cells. As shown in Figure 6B and 6C, the knockdown of ITGAV significantly inhibited the invasion and migration of PANC-1 cells, indicating ITGAV under-expression might inhibit progression of pancreatic cancer cells. Discussion The low survival rate and significant recurrence rate of pancreatic cancer have made the disease well-known 17 . The best course of treatment for localized pancreatic cancer is surgery, although most cases of the disease return following surgery, and most patients pass away within 10 years of being diagnosed 18, 19 . Moreover, the most common event of tumor development in pancreatic cancer is LM, and current treatments have not produced satisfying results 20, 21 . Thus, it is particularly important to explore the mechanism of LM and identify novel biomarker at an early stage of pancreatic cancer. In this study, we identified nine cell clusters in pancreatic cancer PT and LM samples via analyzing single-cell transcriptomic profiles, and found that the proportion of epithelial cells and T cells were higher in PT and LM pancreatic cancer samples. Focused on epithelial cells, we identified totally 15,033 malignant epithelial cells in the PT and LM pancreatic cancer samples. Previous studies demonstrated that compared to normal pancreatic tissue (NT) and PT tissue, tumor cells from liver metastatic lesions (HM) tissue exhibited significantly higher malignant phenotype 22 , which was consistent with our result that the proportion of malignant epithelial cells were higher in LM samples. Compared to non-malignant epithelial cells, malignant epithelial cells were specifically regulated by IL-1, Periostin, and IFN-II signaling pathways. The expression of tumor-derived IL-1α and IL-1β in pancreatic PDAC was associated with a lower survival rate for patients. Additionally, they were a significant part of the inflammatory cascade that triggers tumor-associated macrophages (TAMs) to secrete IL-1β 23 , which was also verified in our research. In this study, the macrophages regulated malignant epithelial cells through the IL-1 pathway. POSTN was found to be strongly expressed in stromal cells adjacent to the pancreatic epithelial cells 24 . Kanno et al. elucidated that POSTN could inhibit growth of PDAC cell 24 . In contrast, some researches considered that POSTN promoted the proliferation and invasiveness of pancreatic cancer 25, 26 . Accordingly, there was a great deal of disagreement regarding POSTN's influence on PDAC cell growth. Further research showed that fibroblasts regulated malignant epithelial cells through the Periostin pathway (POSTN-ITGAV/ITGB5). It has been demonstrated that POSTN can regulate the proliferation of haematopoietic stem cells (HSCs) via interaction with ITGAV, moreover, the interaction between POSTN and ITGAV suppresses the FAK/PI3K/AKT pathway in HSCs, which raises p27Kip1 expression and improves the maintenance of quiescent HSCs 16 . In the present study, POSTN was highly expressed in fibroblast enriched areas, ITGAV was highly expressed in epithelial cell aggregation areas, and the spatial distance between POSTN and ITGAV was close. Therefore, we hypothesized that POSTN might combine with ITGAV to affect the progression of pancreatic cancer via regulating the PI3K signaling pathway, which warrants further exploration in the future studies. Subsequently, we constructed a metastatic risk model (LM-index) using 34 DEGs enriched in the PI3K pathway between malignant and non- malignant epithelial cells. The pancreatic cancer patients with high LM-index had a poor prognosis, indicating this metastatic risk model might effectively predict the prognosis of pancreatic cancer patients. ITGAV is an av protein-encoding gene, the product of ITGAV belongs to the integrin alpha chain family 27 . It has been reported that ITGAV involved in the extracellular matrix (ECM)-receptor interaction pathway 28 . The ITGAV was highly expressed in multiple cancers, such as colorectal cancer 29 , gastric cancer 30 and esophageal adenocarcinoma 31 . The overexpression of ITGAV might be linked to the spread of breast cancer via upregulating PXN 32 . Moreover, ITGAV was differentially expressed among three immune cell infiltration subtypes of metastatic pancreatic neuroendocrine tumors 33 . In pancreatic cancer, ITGAV could affect the prognosis and clinicopathological factors of tumor metastasis, and the patients with high ITGAV expression exhibited inferior prognosis and recurrence rate compared to patients with low ITGAV expression 34 . In addition, by upregulating the expression of ITGAV, A1B1 can promote ECM signal transduction, thereby accelerating progression of PDAC 35 . Correspondingly, in PDAC cells, ITGAV knockdown significantly decreased peritoneal carcinomatosis, spontaneous lung metastasis, and primary tumor development 36 . In our study, we also found that the knockdown of ITGAV significantly inhibited the invasion and migration of PANC-1 cells, indicating ITGAV under-expression might inhibit progression of pancreatic cancer cells. However, the mechanism of ITGAV affecting the progression of pancreatic cancer needs to be further explored in future studies. In conclusion, this study clarified possible cellular origins and drivers of LM pancreatic cancer based on the single-cell transcriptomic profiles, and discovered that ITGAV mRNA and protein were expressed in pancreatic cancer tissue and cells, and demonstrated that knockdown of ITGAV inhibited the progression of pancreatic cancer. Our results provide more reference information for understanding the mechanism of ITGAV in LM pancreatic cancer patients. Abbreviations LM Liver metastasis PT Primary tumor CNV Copy number variation TME Tumor microenvironment PDAC Pancreatic ductal adenocarcinoma EMT Epithelial-mesenchymal transition DEGs Differentially expressed genes GO Gene Ontology KEGG Kyoto Encyclopedia of Genes and Genome GSVA Gene set variation analysis BEAM Branch expression analysis modeling OCLR One class logistic regression FBS Fetal bovine serum siRNA Small interfering RNA NC Negative control HSCs Haematopoietic stem cells Declarations Funding Information This study did not receive any specific funding. Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Data Availability Statement The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material .All data are available from the TCGA database (https://xenabrowser.net/datapages/) and GEO (https://ncbi.nlm.nih.gov/geo/) database within the article. GEO database under accession number GSE154778, GSE154778andGSE202740. Ethics Statement Approval of the research protocol by an Institutional Review Board: The use of the human tissue microarrays was approved by the Ethics Committee of Union Hospital of Tongji Medical College of Huazhong University of Science and Technology. Informed Consent: N/A. Registry and the Registration No. of the study/trial: N/A Animal Studies: N/A. Author Contributions JC and TP designed and supervised the entire work. JC, YG, TC, JX and ZK carried out the bioinformatics analysis. YL, XL and TY performed the experiments. JC, YG and JZ analysed the data and prepared the figures. JC and TP prepared and wrote the paper. All authors contributed to the article and approved the submitted version. References Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA: a cancer journal for clinicians . 2022; 72: 7-33. Ritzkowsky A. [Accidental needle injuries of medical personnel]. Dtsch Med Wochenschr . 1994; 119: 1563-1564. Xia C, Dong X, Li H, et al. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chin Med J (Engl) . 2022; 135: 584-590. Moore A, Donahue T. Pancreatic Cancer. Jama . 2019; 322: 1426. Kolbeinsson HM, Chandana S, Wright GP, Chung M. Pancreatic Cancer: A Review of Current Treatment and Novel Therapies. J Invest Surg . 2023; 36: 2129884. Torphy RJ, Fujiwara Y, Schulick RD. Pancreatic cancer treatment: better, but a long way to go. Surg Today . 2020; 50: 1117-1125. Mizrahi JD, Surana R, Valle JW, Shroff RT. Pancreatic cancer. Lancet (London, England) . 2020; 395: 2008-2020. Li CG, Zhou ZP, Jia YZ, Tan XL, Song YY. Radioactive (125)I seed implantation for pancreatic cancer with unexpected liver metastasis: A preliminary experience with 26 patients. World J Clin Cases . 2021; 9: 792-800. Ducreux M, Cuhna AS, Caramella C, et al. Cancer of the pancreas: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Annals of oncology : official journal of the European Society for Medical Oncology . 2015; 26 Suppl 5: v56-68. Gai B, Zhang F. Chinese expert consensus on radioactive (125)I seeds interstitial implantation brachytherapy for pancreatic cancer. Journal of cancer research and therapeutics . 2018; 14: 1455-1462. Chen K, Wang Q, Li M, et al. Single-cell RNA-seq reveals dynamic change in tumor microenvironment during pancreatic ductal adenocarcinoma malignant progression. EBioMedicine . 2021; 66: 103315. Sang W, Zhou Y, Chen H, et al. Receptor-interacting protein kinase 2 is an immunotherapy target in pancreatic cancer. Cancer discovery . 2023. Zhao L, Zhao H, Yan H. Gene expression profiling of 1200 pancreatic ductal adenocarcinoma reveals novel subtypes. BMC cancer . 2018; 18: 603. Peng J, Sun BF, Chen CY, et al. Single-cell RNA-seq highlights intra-tumoral heterogeneity and malignant progression in pancreatic ductal adenocarcinoma. Cell research . 2019; 29: 725-738. Ligorio M, Sil S, Malagon-Lopez J, et al. Stromal Microenvironment Shapes the Intratumoral Architecture of Pancreatic Cancer. Cell . 2019; 178: 160-175 e127. Khurana S, Schouteden S, Manesia JK, et al. Outside-in integrin signalling regulates haematopoietic stem cell function via Periostin-Itgav axis. Nature communications . 2016; 7: 13500. Chen W, Zheng R, Baade PD, et al. Cancer statistics in China, 2015. CA: a cancer journal for clinicians . 2016; 66: 115-132. Kim J. Cell Dissemination in Pancreatic Cancer. Cells . 2022; 11. Karunakaran M, Barreto SG. Surgery for pancreatic cancer: current controversies and challenges. Future Oncol . 2021; 17: 5135-5162. He H, Zhang S, Yang H, et al. Identification of Genes Associated with Liver Metastasis in Pancreatic Cancer Reveals PCSK6 as a Crucial Mediator. Cancers . 2022; 15. Nagai M, Wright MJ, Ding D, et al. Oncologic resection of pancreatic cancer with isolated liver metastasis: Favorable outcomes in select patients. J Hepatobiliary Pancreat Sci . 2023; 30: 1025-1035. Zhang S, Fang W, Zhou S, et al. Single cell transcriptomic analyses implicate an immunosuppressive tumor microenvironment in pancreatic cancer liver metastasis. Nature communications . 2023; 14: 5123. Brunetto E, De Monte L, Balzano G, et al. The IL-1/IL-1 receptor axis and tumor cell released inflammasome adaptor ASC are key regulators of TSLP secretion by cancer associated fibroblasts in pancreatic cancer. J Immunother Cancer . 2019; 7: 45. Kanno A, Satoh K, Masamune A, et al. Periostin, secreted from stromal cells, has biphasic effect on cell migration and correlates with the epithelial to mesenchymal transition of human pancreatic cancer cells. International journal of cancer . 2008; 122: 2707-2718. Erkan M, Kleeff J, Gorbachevski A, et al. Periostin creates a tumor-supportive microenvironment in the pancreas by sustaining fibrogenic stellate cell activity. Gastroenterology . 2007; 132: 1447-1464. Ben QW, Jin XL, Liu J, Cai X, Yuan F, Yuan YZ. Periostin, a matrix specific protein, is associated with proliferation and invasion of pancreatic cancer. Oncol Rep . 2011; 25: 709-716. Chen ZZ, Wang WP, Xue HM, Liang Y. The lncRNA-miRNA-integrin alpha V ceRNA network can affect the occurrence and prognosis of gastric cancer. Int J Clin Exp Pathol . 2022; 15: 388-402. He B, He Y, Shi W, et al. Bioinformatics analysis of gene expression alterations in microRNA‑122 knockout mice with hepatocellular carcinoma. Mol Med Rep . 2017; 15: 3681-3689. Waisberg J, De Souza Viana L, Affonso Junior RJ, et al. Overexpression of the ITGAV gene is associated with progression and spread of colorectal cancer. Anticancer research . 2014; 34: 5599-5607. Zhang W, Chen Y, Qiao Z, Liu Y. Overexpression of Integrin alpha V (ITGAV) in gastric cancer and its prognostic significance. Asian J Surg . 2023; 46: 5863-5864. Loeser H, Scholz M, Fuchs H, et al. Integrin alpha V (ITGAV) expression in esophageal adenocarcinoma is associated with shortened overall-survival. Scientific reports . 2020; 10: 18411. Cheuk IW, Siu MT, Ho JC, Chen J, Shin VY, Kwong A. ITGAV targeting as a therapeutic approach for treatment of metastatic breast cancer. Am J Cancer Res . 2020; 10: 211-223. Meng D, Zhao L, Liu J, Ge C, Zhang C. Identification of the Immune Subtypes for the Prediction of Metastasis in Pancreatic Neuroendocrine Tumors. Neuroendocrinology . 2023; 113: 719-735. Iwatate Y, Yokota H, Hoshino I, et al. Machine learning with imaging features to predict the expression of ITGAV, which is a poor prognostic factor derived from transcriptome analysis in pancreatic cancer. Int J Oncol . 2022; 60. Li L, Bao J, Wang H, et al. Upregulation of amplified in breast cancer 1 contributes to pancreatic ductal adenocarcinoma progression and vulnerability to blockage of hedgehog activation. Theranostics . 2021; 11: 1672-1689. Kemper M, Schiecke A, Maar H, et al. Integrin alpha-V is an important driver in pancreatic adenocarcinoma progression. Journal of experimental & clinical cancer research : CR . 2021; 40: 214. Additional Declarations No competing interests reported. Supplementary Files FigureS1.pdf TableS1S10.xlsx ListofSupportingInformation.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4128922","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":290761204,"identity":"eab12e57-d3df-4336-b59a-c4fe2a59109b","order_by":0,"name":"Jing Cui","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/UlEQVRIiWNgGAWjYFACxgaDBAYGHn725oOPoUIGRGmRkew5lmwM5EoQoQUCbAxu+KhJE6WFf0ZyQ8HDHbU8DDd42KoL2+rqGNibt0kw1NzBqUXiRmKDQeKZ4zyMs3uP3Z7ZdliCgedYmQTDsWc4tRhIgLS0HeNhljmXdpu37YAEg0SOmQRjw2HCWtiAKot52+okGOTfEKWlhocHqIWZt40ZaAsPfi0SZx6CtBzgkeA5lizNc+6wZBtPWrFFwjHcWvjb058Z/myrs7c/3nzwM09ZHT8/++GNNz7U4NYCBGzAaEBSwAYiEvBpYGBgfsDAUIdfySgYBaNgFIxsAAAzBFCP9omc0AAAAABJRU5ErkJggg==","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Jing","middleName":"","lastName":"Cui","suffix":""},{"id":290761205,"identity":"f9dc3e5d-742d-4c8d-b9b1-3186e442ab8f","order_by":1,"name":"Yao Guo","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yao","middleName":"","lastName":"Guo","suffix":""},{"id":290761206,"identity":"17921fa3-4ab6-42dc-8717-9b1e45ad96f3","order_by":2,"name":"Taoyu Chen","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Taoyu","middleName":"","lastName":"Chen","suffix":""},{"id":290761207,"identity":"c71a1717-fa94-43de-a06a-03526f5c64c6","order_by":3,"name":"Yan Ling","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Ling","suffix":""},{"id":290761208,"identity":"4a3768ec-90d8-4691-a64c-bc1c7ed05548","order_by":4,"name":"Xueyi Liang","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xueyi","middleName":"","lastName":"Liang","suffix":""},{"id":290761209,"identity":"36709b43-b029-44ed-86ea-cff5167d84dc","order_by":5,"name":"Jingyuan Zhao","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Jingyuan","middleName":"","lastName":"Zhao","suffix":""},{"id":290761210,"identity":"62f7351f-83fa-489e-85c3-27aa231a31c2","order_by":6,"name":"Jiongxin Xiong","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Jiongxin","middleName":"","lastName":"Xiong","suffix":""},{"id":290761211,"identity":"67125f67-fa55-45d9-9550-8feefece8aef","order_by":7,"name":"Tao Yin","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Yin","suffix":""},{"id":290761212,"identity":"a5f9837e-bfe5-4be1-8c27-3c9323baf9f3","order_by":8,"name":"Zunxiang Ke","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Zunxiang","middleName":"","lastName":"Ke","suffix":""},{"id":290761213,"identity":"47d87fb0-e2e3-4b36-abd8-614389afdd9a","order_by":9,"name":"Tao Peng","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Peng","suffix":""}],"badges":[],"createdAt":"2024-03-19 09:14:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4128922/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4128922/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":54928856,"identity":"5d7a814e-f2ab-4a57-87fc-bcd776a894c9","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":682058,"visible":true,"origin":"","legend":"\u003cp\u003eCharacterization of discrete cellular compositions in pancreatic cancer via meticulous single-cell analysis. (A) the result of cell annotation in PT samples. (B) the expression of marker genes in nine cell clusters in PT samples. (C) the number of cells in the nine clusters in PT samples. (D) the result of cell annotation in LM pancreatic cancer samples. (E) the expression of marker genes in nine cell clusters in LM pancreatic cancer samples. (F) the number of cells in the nine clusters in LM pancreatic cancer samples. (G) the proportion of nine cell clusters in PT and LM pancreatic cancer samples. (H) GSVA enrichment results of various cells in PT samples. (I) GSVA enrichment results of various cells in LM pancreatic cancer samples.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/15d93201f4dac59e10ee17c6.png"},{"id":54928854,"identity":"817a18d7-35f6-400e-9ff7-e70ea024cd0a","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1047010,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of malignant epithelial cells in pancreatic cancer. (A) the copy number variation (CNV) of epithelial cells in PT samples. (B) the CNV of epithelial cells in LM pancreatic cancer samples. (C) the proportion of malignant epithelial cells in PT and LM pancreatic cancer samples. (D) evolutionary tree of all malignant cells in PT samples. (E) evolutionary tree of all malignant cells in LM pancreatic cancer samples (Subclones with \u0026lt; 5% cell number were not included). (F) chromosome gene upset map of all ma-lignant cells in PT samples. (G) chromosome gene upset map of all malignant cells in LM pancreatic cancer samples. (H) CNV genes shared by all malignant cells in PT and LM pancreatic cancer samples.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/9891b8c75d18a900472ccb83.png"},{"id":54928857,"identity":"418a2dfa-2589-4763-aaaa-fb445e72d01f","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":599363,"visible":true,"origin":"","legend":"\u003cp\u003eThe results of intercellular interaction. (A) the number of non-malignant epi-thelial cells and other cell-cell interactions in PT samples. (B) the number of malignant epithelial cells and other cell-cell interactions in LM samples. (C) interaction strength between non-malignant epithelial cells and other cells in PT samples. (D) interaction strength between malignant epithelial cells and other cells in LM samples. (E) differential cell-cell interaction signaling pathways between malignant epithelial cells and non-malignant epithelial cells in PT samples. Cell to cell interactions of IL1 (F), PERI-OSTIN (G) and IFN-II (H) pathway. (I) receptors and ligands in monocytes, NK cells, fi-broblasts, macrophages and epithelial (malignant/non-malignant) cells.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/0a04f09e9453f26c7312cc9e.png"},{"id":54929304,"identity":"2683332d-de33-4338-a92d-6aa5554b76a0","added_by":"auto","created_at":"2024-04-18 17:48:31","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1989951,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distance between POSTN and ITGAV in pancreatic cancer. (A) annotation results of spatial sequencing of pancreatic cancer. (B) expression of POSTN and ITGAV in pancreatic cancer spatial data. (C) Expression of POSTN (green) and ITGAV (red) in tissue sections of pancreatic cancer patients.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/b07e1556b1eff14d3665a06e.png"},{"id":54928859,"identity":"ddbb93e9-306c-4313-8a3a-24d9f56b757c","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1095919,"visible":true,"origin":"","legend":"\u003cp\u003eITGAV was highly expressed in pancreatic cancer. (A) the expression of ITGAV in tissue sections of pancreatic cancer patients. (B) the expression of ITGAV mRNA in PANC-1 and HPDE6-C7 cell lines. * p \u0026lt; 0.05, ** p \u0026lt;0.01, **** p \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/eb661ce669ffbbc79ea4540c.png"},{"id":54928861,"identity":"8b50f0c3-503e-4307-b3db-c87d588b8138","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1040951,"visible":true,"origin":"","legend":"\u003cp\u003eKnockdown of ITGAV inhibited the invasion and migration of pancreatic cancer cells. (A) the expression of ITGAV in siITGAV-1, siITGAV-2, siITGAV-3 groups. (B) the invasion capacity of siITGAV-2, siITGAV-3 groups. (C) the migration capacity of si-ITGAV-2, siITGAV-3 groups. (D) mechanism of ITGAV in the progression of pancreatic cancer. ** p \u0026lt;0.01, *** p \u0026lt; 0.001, **** p \u0026lt; 0.0001.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/64cbd0011943ffc727d8d56f.png"},{"id":60787139,"identity":"98565fbb-3cb6-4c7d-a364-244fadefe762","added_by":"auto","created_at":"2024-07-22 06:05:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7277907,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/3285e252-e96c-4612-b88c-13b72716bbcd.pdf"},{"id":54928853,"identity":"93b0f81e-f5ed-4d61-9d1a-9a476f8f86a7","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1201002,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/78f61a20c7da5249f254a65a.pdf"},{"id":54928862,"identity":"13788f85-b207-4795-96f5-17a6d141451d","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3786268,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1S10.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/7ae5219085dda770f83c6a11.xlsx"},{"id":54928855,"identity":"478a8731-8f76-46cd-947d-829518d4a68a","added_by":"auto","created_at":"2024-04-18 17:40:31","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":15294,"visible":true,"origin":"","legend":"","description":"","filename":"ListofSupportingInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-4128922/v1/aa59f0c600eee219861f4bea.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Single-cell and spatial transcriptomic analyses reveal the cellular origins and drivers of liver metastasis from pancreatic cancer","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003ePancreatic cancer is currently the seventh leading cause of cancer-related death worldwide, and it is also the malignant tumor with the mortality closest to the incidence, which is characterized by high recurrence rate and low survival rate\u0026nbsp;\u003csup\u003e1, 2\u003c/sup\u003e.\u0026nbsp;According to statistics, in 2022, there were 134374 new patients with pancreatic cancer and 131203 deaths from pancreatic cancer in China\u0026nbsp;\u003csup\u003e3\u003c/sup\u003e. The overall population incidence ranked eighth in the tumor category, and the mortality ranked sixth in the tumor category. Among them, the male incidence ranked fifth and the mortality ranked third, both higher than the female, showing certain gender differences\u0026nbsp;\u003csup\u003e3\u003c/sup\u003e. Pancreatic cancer risk factors include smoking, chronic pancreatitis, obesity, long-term diabetes, a significant family history of pancreatic cancer, and diets heavy in red and processed meats\u0026nbsp;\u003csup\u003e4\u003c/sup\u003e. Currently, surgery is the most effective treatment for localized pancreatic cancer\u0026nbsp;\u003csup\u003e5, 6\u003c/sup\u003e, but, less than 20% of patients are suitable for curative resection at the first diagnosis, and chemotherapy remains the cornerstone of treatment for both unresectable advanced and metastatic pancreatic cancer\u0026nbsp;\u003csup\u003e7\u003c/sup\u003e. It has been reported that the most common metastatic site of pancreatic cancer is the liver, and liver metastasis (LM) occurs in about 70% of pancreatic cancer patients\u0026nbsp;\u003csup\u003e8, 9\u003c/sup\u003e, and the average survival time of pancreatic cancer patients with liver metastases is about 3 to 6 months\u0026nbsp;\u003csup\u003e10\u003c/sup\u003e. Accordingly, it is particularly important to characterize the molecular characteristics of LM of pancreatic cancer, which may help to find a new scheme for the diagnosis of early pancreatic cancer.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn recent years, the application of single-cell sequencing data has revealed dynamic changes in the tumor microenvironment (TME) during malignant progression of pancreatic cancer\u0026nbsp;\u003csup\u003e11\u003c/sup\u003e. Sang et al. have found that CD8+T cells are the immune cell type most severely affected by ripk2 deficiency in the TME of pancreatic ductal adenocarcinoma (PDAC)\u0026nbsp;\u003csup\u003e12\u003c/sup\u003e. Zhao et al. have collected the whole transcriptome data of 1200 PDAC patients. Through retrospective meta-analysis, they divided PDAC into six molecular subtypes, each of which has its characteristic gene expression profile and shows different clinical characteristics, so as to better stratify PDAC patients and provide personalized treatment\u0026nbsp;\u003csup\u003e13\u003c/sup\u003e. By analyzing the single-cell RNA-seq data of 24 primary PDAC and 11 normal pancreases, Peng et al. have found that the proliferation of subpopulations with high expression of cancer cells associated with malignant behavior is often accompanied by the loss of activated T cells, which predicts a poor prognosis\u0026nbsp;\u003csup\u003e14\u003c/sup\u003e. Moreover, Ligorio et al. have detected notable changes in single-cell populations towards proliferative (PRO) and invasive epithelial-to-mesenchymal transition (EMT) phenotypes in pancreatic cancer, which are associated with signal transducer and activator of transcription 3 (STAT3) and mitogen-activated protein kinase (MAPK) signaling by using single-cell RNA sequencing technology and RNA in situ hybridization technologies\u0026nbsp;\u003csup\u003e15\u003c/sup\u003e. Therefore, analyzing the single-cell RNA-seq data of pancreatic cancer will help determine the cell subtypes and explore the therapeutic strategies for pancreatic cancer.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study focused on LM from pancreatic cancer. By analyzing single-cell transcriptome data from PT and LM pancreatic cancer samples, we identified nine cell clusters and malignant epithelial cells. Moreover, we constructed a metastasis-related prognosis signature in pancreatic cancer, and analyzed spatial distance between POSTN and ITGAV in pancreatic cancer based on the spatial transcriptomic profiles. Finally, we explored the role of \u003cem\u003eITGAV\u003c/em\u003e in the progression of pancreatic cancer. Our findings are expected to provide further insights into the development of therapeutic strategies for pancreatic cancer.\u003c/p\u003e"},{"header":"2 Materials And Methods","content":"\u003cp\u003e\u003cstrong\u003e2.1 Data collection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSingle cell sequencing data of pancreatic cancer were downloaded from Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/, ID: GSE154778 and GSE154778) database. GSE156405 contained 5 primary tumor (PT) and 1 liver metastatic (LM) pancreatic cancer samples, and GSE154778 included 10 PT samples and 5 LM samples (Table S1). The transcriptome data of 183 pancreatic cancer samples were retrieved from The Cancer Genome Atlas (TCGA, https://tcga-data.nci.nih.gov/tcga/) database. Among which, totally 126 samples had complete survival information (Table S2). Furthermore, spatial transcriptome data of PT samples (ST_PDAC-1 (GSM6132061)) by high throughput sequencing were extracted from GSE202740 dataset.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2 Single cell data quality control and cell annotation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCell Ranger (v6.1.2) was utilized to align the data to the human genome (GRCh37). Seurat v4.1.1 was used to process single-cell data, and cells with mitochondrial content higher than 20%, hemoglobin content higher than 5%, and expression genes less than 200 were filtered. Seurat was used for data normalization, cell clustering and dimension reduction. The \u0026quot;FindVariableFeatures\u0026quot; function was used to select 2000 highly variable genes from the corrected expression matrix, and then the \u0026quot;RunPCA\u0026quot; function was used for principal component analysis, retaining the top 20 principal components for further analysis. Batch effects were corrected by \u0026quot;RunHarmony\u0026quot; of R package harmony. The \u0026quot;FindClusters\u0026quot; function was used for cell clustering (resolution 0.6), and the \u0026quot;RunUMAP\u0026quot; function was used for nonlinear dimensionality reduction. Cell grouping was annotated based on cellmark2.0 database and common mark genes of cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.3 Differential gene and functional enrichment analyses\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe \u0026ldquo;Findmarkers\u0026rdquo; function from the Seurat package was utilized to performed differential gene analysis. The differentially expressed genes (DEGs) among between two groups were screened by p.adj \u0026lt; 0.05. The R package \u0026quot;ClusterProfiler\u0026quot; was used for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome (KEGG) enrichment analysis. The significantly enriched pathways were screened using p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.4 Gene set variation analysis (GSVA)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Hallmark gene set was downloaded for GSVA from the MSigDB(v2023.1) (https://www.gsea-msigdb.org) database to investigate the difference of the biological function between two groups using R package. The difference analysis among groups were conducted using the \u0026quot;limma\u0026quot; package. The differentially enriched pathways between two groups were identified based on |t| \u0026gt; 2 and p \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.5 Copy number variation (CNV) analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on single-cell gene expression and chromosome sequencing data, the inferCNV (v1.14.0) package in R was used to distinguish malignant epithelial cells from non-malignant epithelial cells. The inferCNV analysis was performed with the following settings: cutoff=0.1, cluster-by-groups=TRUE, tumor subcluster-partition-method = \u0026quot;random-trees\u0026quot;, and hidden markov model (HMM) = TRUE. To reduce false positive calls in CNV inference, the default Bayesian latent mixture model was applied to determine the posterior probability of changes in each cell, with a default threshold of 0.5. Furthermore, all gene CNV scores of ECs and reference cells (T cells) were hierarchically clustered using the k-means algorithm. The malignant epithelial cells were identified according to the CNV score. If the score was greater than the 95th percentile of the reference cell, it is malignant epithelial cells. The others were non-malignant epithelial cells.\u003c/p\u003e\n\u003cp\u003eTo illustrate tumor clonality and evolution, the \u0026quot;sub cluster\u0026quot; pattern was further applied to divide malignant cells into eight clusters, and HMM generated different CNV patterns. Each CNV was annotated as a gain or loss at the p or q arm level according to UCSC chromosome band information. Subclones containing the same arm level CNVs were collapsed to construct an evolutionary tree. The phylogenetic tree was drawn using Uphyloplot2 software to represent the subclonal CNV structure.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.6 Inference of developmental trajectory\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Monocle (v2.28.0) was used to construct pesudotime trajectories based on gene expression profiles of malignant epithelial cells. After dimensionality reduction and cell sorting, all malignant epithelial cells were projected and sorted into trajectories with different branches, and cells within the same branch were considered to have the same cell state. Branch expression analysis modeling (BEAM) was further performed to identify genes with branch dependent expression patterns. These branch dependent genes can help us to explore the mechanism of cell fate determination.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.7 Cell communication analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCellchat (v1.5.0) package was used to predict and visualize biologically meaningful intercellular communication. For each individual dataset, first extracted the expression matrix and metadata from the Seurat object, and then used the \u0026quot;createCellChat\u0026quot; function to generate a CellChat object. After calculating the highly variable genes and pathways, the \u0026quot;computeCommunProb\u0026quot; function was used to infer the intercellular communication probability. The results were shown using a series of visualization functions provided by CellChat, such as \u0026quot;netVisual-bubble\u0026quot; presented the dot diagram of signaling pathways emitted from cells. The functions \u0026quot;compareInteractions\u0026quot;, \u0026quot;netVisual-diffinteraction\u0026quot;, \u0026quot;netVisual-heatmap\u0026quot; of the CellChat package were used to compare the relative number or interaction strength of various cell subsets between two groups.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.8 LM-index and overall survival analysis\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the risk genes, the gelnet (v1.2.1) package in R was applied to construct a LM-index by one class logistic regression (OCLR) algorithm to represent the risk of LM pancreatic cancer samples. The patients were divided into high and low LM-index groups using the R package \u0026quot;survminer\u0026quot;. Kaplan Meier (K-M) survival analysis was used to evaluate the overall survival (OS) of patients with high and low LM-index, and two-sided log rank test was used for comparison.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.9 Spatial data analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRaw data was quality controlled using fastp, and then data alignment and count were performed using SpaceRanger (v2.1.0). Seurat package (v4.4.0) was used for data filtering (spot with more than 200 retained genes, less than 20% mitochondria, and less than 5% hemoglobin), spot normalization and regression. Finally, Seurat\u0026apos;s FindTransferAnchors and transferdata functions were employed to map single-cell data and achieve annotation of spatial data.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.10 Immunofluorescence staining\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePancreatic cancer tissue microarray samples were purchased from bioaitech (Xi\u0026rsquo;an,China). The use of the human tissue microarrays was approved by the Ethics Committee of Union Hospital of Tongji Medical College of Huazhong University of Science and Technology. The in-formation of patients was shown in Table S3. ITGAV (Proteintech, #27096-1-ap) and postn (Proteintech, #66491-1-Ig) were used as primary antibodies.\u003c/p\u003e\n\u003cp\u003eImmunofluorescence staining was performed according to the instructions of alphatsa multiplex IHC Kit (AXT37100031, Alphaxbio). In brief, the tissue chip was dewaxed and hydrated through a series of xylene and alcohol washes, and then antigen repair and sealing were performed. The sections were blocked and incubated with primary antibody and secondary antibody, and fluorescent staining was performed. Finally, the nuclei were counterstained with DAPI for 5 min and enclosed in Mounting Medium. ZEN (v3.1) software was used for film reading.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.11 Cell collection and culture\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHuman pancreatic cancer cell line PANC-1 and human normal pancreatic ductal epithelial cell line HPDE6-C7 were purchased from BeNa culture collection (Beijing, China). All cells were cultured in RPMI-1640 complete medium (Gibco) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. The cells were placed in a humidified incubator maintained at 37 \u0026deg;C with 5% CO\u003csub\u003e2\u003c/sub\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.12 qRT-PCR\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal of RNA from cells was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA), and the quality and concentration of RNA were evaluated using UV spectrophotometer, and then reverse transcription was performed using Transcriptor First Strand cDNA Synthesis Kit (GenStar, Beijing, China). Furthermore, qPCR detection was performed using the LightCycler 480 fluorescence quantification system (Roche, Basel, Switzerland). The reference gene was GADPH, and the primer sequences were listed in Table S4. The mRNA expression levels were calculated using the 2-\u0026Delta;\u0026Delta;CT method (three repeats).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.13 Cell transfection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe small interfering RNA (siRNA) was used to knock down ITGVA in PANC-1 cells. The ITGAV siRNA sequence was presented in Table S5. The cells were seeded onto 6-well plates and transfected once they reached approximately 50% confluence. The transfection involved using a negative control (NC) and si-ITGAV to transfect the cells, respectively. All transfections were carried out with Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). The knockdown efficiency of ITGAV was detected by qRT-PCR.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.14 Cell scratch assay\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 7\u0026times;10\u003csup\u003e5\u003c/sup\u003e cells were seeded in each well of a 6-well and cultured for 24 hours. A line was drawn in the center of the well using a 10 \u0026mu;L pipette tip. After two washes with PBS, the cells were cultured in a 37 \u0026deg;C incubator for 24 hours. Subsequently, the wounds were photographed using a microscope at various time intervals. The distances of the wounds were measured using Fiji (ImageJ).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.15 Invasion and migration assays\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe migration and invasion abilities of cells were analyzed using a polycarbonate membrane with an 8 \u0026mu;m pore size in a 24-well Transwell chamber (Coring, NY, USA). The upper chamber added 1\u0026times;10\u003csup\u003e4\u003c/sup\u003e cells in serum-free medium containing 0.1% BSA. Moreover, the lower chamber was supplemented with medium containing 0.1% BSA and EGF (50 ng/ml, MCE, NJ, USA). After incubation for 24 h, the cells in the upper chamber were completely transferred to the lower membrane. The polycarbonate membrane was fixed and stained with Giemsa solution (Solarbio, Beijing, China), and photographed with a microscope.\u003c/p\u003e"},{"header":"3 Results","content":"\u003cp\u003e\u003cstrong\u003e3.1 Characterization of discrete cellular compositions in pancreatic cancer via meticulous single-cell analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFirstly, we extracted totally 35682 cells from PT and LM pancreatic cancer samples, including 17961 PT cells and 17721 LM pancreatic cancer cells. Subsequently, these 35682 cells were clustered into 9 clusters by cell clustering: Epithelial cells (KRT19), T cells (CD3D), macrophage (SPP1), fibroblast (COL3A1), monocyte (S100A9), natural killer (NK: NKG7), B cells (CD79A), endothelial (COL4A1/CLDN5), dendritic cell (DC: STMN1/JCHAIN) (Figure 1A-1F). The proportion of epithelial cells and T cells were higher in PT and LM pancreatic cancer samples. Compared to LM samples, the proportion of fibroblast and macrophage were increased in PT samples (Figure 1 G).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo elucidate the involvement of cell clusters in the metastatic process of pancreatic cancer, we conducted a GSVA enrichment analysis of cell clusters in the PT and LM samples. In the PT samples, the glycolysis and metabolic related pathways, such as Cholesterol homeostasis, Bile acid metabolism were significantly activated in the epithelial cell cluster, and epithelial-mesenchymal transition (EMT) and angiogenesis pathways were remarkably activated in the fibroblast cluster (Figure 1H). Compared to PT samples, Angiogenesis, Hedgehog signaling, E2F targets pathways were significantly activated in LM samples in the epithelial cell cluster, and oxidative phosphorylation and glycolysis pathways were observably activated in LM samples in the fibroblast cluster (Figure 1I).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2 Identification of malignant epithelial cells in pancreatic cancer\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine the clonal structure and cellular origin of malignant epithelial cell, we calculated the CNV and clonality of epithelial cells in PT and LM samples via inferCNV algorithm. Compared to the reference cells, a total of 6396 cells with higher CNV score were considered malignant epithelial cells among 6903 epithelial cells of PT samples (Figure 2A). Meanwhile, among the 8772 epithelial cells of LM samples, totally 8637 cells were considered malignant epithelial cells (Figure 2B). The proportion of malignant epithelial cells were higher in LM samples (Figure 2C, LM vs. PT).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFurthermore, we utilized phylogenetic trees to illustrate the clonality of tumors and the progression of malignant cells from PT to LM pancreatic cancer samples. In all malignant epithelial cells of PT and LM samples, gain of 1q and loss of 11q were observed (Figure 2D, 2E). Totally 700 CNV changed genes were shared by 8 subclonal cell populations of PT samples (Figure 2F, Table S6), and 225 CNV changed genes were shared by all subclonal cell populations of LM samples (Figure 2G, Table S7). Moreover, we found that when pancreatic cancer tumors metastasized to the liver, copy number changes in 27 genes were retained, while LM samples also produced additional changes in 198 genes (Figure 2H). These findings indicated that malignant epithelial cells in LM might come from PT subclones with 27 common gene CNV, and malignant cells also generated more new CNV subclones in the process of transferring to liver tissue.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.3 The results of intercellular interaction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNext, we used Cellchat to analyze the cell-cell communication among malignant epithelial cells, non-malignant epithelial cells and other cells in PT samples, and the results showed that the interaction between other cells and malignant epithelial cells increased (Figure 3A-3D, Table S8-9). Compared to non-malignant epithelial cells, malignant epithelial cells were specifically regulated by IL-1, Periostin, and IFN-II signaling pathways (Figure 3E). Monocytes and macrophages regulated malignant epithelial cells through the IL-1 pathway (IL-1B-IL1R2), fibroblasts regulated malignant epithelial cells through the Periostin pathway (POSTN-ITGAV/ITGB5), and NK cells regulated malignant epithelial cells through the IFN-II pathway (IFNG-IFNGR1/IFNGR2) (Figure 3F-3I).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.4 Construction of metastatic risk model in pancreatic cancer\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGO enrichment analysis showed that the DEGs between malignant and non- malignant epithelial cells were highly enriched in cadherin binding, actin binding, cell substrate junction, focal adhesion, and actin filament\u0026nbsp;organization (Figure S1A). KEGG enrichment analysis showed that these DEGs were highly enriched in Tight junction, PI3K-Akt signaling pathway and Focal adhension (Figure S1B). After combined with ITGAV, POSTN can trigger changes in the intracellular PI3K signaling pathway, which leads to cancer cell proliferation and invasion\u0026nbsp;\u003csup\u003e16\u003c/sup\u003e. Thus, we defined 34 genes enriched in PI3K pathway by malignant cells (Table S10) as risk genes and calculated liver metastatic (LM) index. According to the LM-index, the pancreatic cancer patients were divided into high and low LM-index groups. The pancreatic cancer patients with high-index exhibited a poor prognosis (Figure S1C, p =0.046). SCENIC analysis showed that ELF3, EHF, YY1, KLF2 and CREB3L1 were enriched in malignant cells (Figure S1D).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.5 Spatial distance between POSTN and ITGAV in pancreatic cancer\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe mapped the single-cell data annotation results to the pancreatic cancer spatial sequencing samples and found that the samples contained fibroblasts, epithelial cells, endothelial cells and macrophages, and fibroblasts were adjacent to epithelial cells (Figure 4A). The expressions of POSTN and ITGAV in tissue sections were shown in Figure 4B. POSTN was highly expressed in fibroblast enriched areas, and ITGAV was highly expressed in epithelial cell aggregation areas. Subsequently, we used the cancer tissue sections of 6 pancreatic cancer patients for fluorescence staining analysis, and the results showed that the spatial distance between POSTN and ITGAV was close (Figure 4C).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6 Knockdown of ITGAV inhibited the invasion and migration of pancreatic cancer cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe found that ITGAV was expressed in tissue sections of six pancreatic cancer patients (Figure 5A), and the expression of ITGAV was significantly upregulated in PANC-1 compared with human normal pancreatic ductal epithelial cell line HPDE6-C7 (Figure 5B). In addition, to explore the effect of ITGAV in the progression of pancreatic cancer, we constructed ITGAV knockdown (si-ITGAV-1, siI-TGAV-2, si-ITGAV-3) PANC-1 cells, and discovered that ITGAV expression was significantly reduced in si-ITGAV-1, si-ITGAV-2, si-ITGAV-3 groups (Figure 6A). Moreover, we analyzed the ITGAV knockdown\u0026rsquo;s impacts on invasion and migration of PANC-1 cells. As shown in Figure 6B and 6C, the knockdown of ITGAV significantly inhibited the invasion and migration of PANC-1 cells, indicating ITGAV under-expression might inhibit progression of pancreatic cancer cells.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe low survival rate and significant recurrence rate of pancreatic cancer have made the disease well-known\u0026nbsp;\u003csup\u003e17\u003c/sup\u003e. The best course of treatment for localized pancreatic cancer is surgery, although most cases of the disease return following surgery, and most patients pass away within 10 years of being diagnosed\u0026nbsp;\u003csup\u003e18, 19\u003c/sup\u003e. Moreover, the most common event of tumor development in pancreatic cancer is LM, and current treatments have not produced satisfying results\u0026nbsp;\u003csup\u003e20, 21\u003c/sup\u003e. Thus, it is particularly important to explore the mechanism of LM and identify novel biomarker at an early stage of pancreatic cancer. In this study, we identified nine cell clusters in pancreatic cancer PT and LM samples via analyzing single-cell transcriptomic profiles, and found that the proportion of epithelial cells and T cells were higher in PT and LM pancreatic cancer samples.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFocused on epithelial cells, we identified totally 15,033 malignant epithelial cells in the PT and LM pancreatic cancer samples. Previous studies demonstrated that compared to normal pancreatic tissue (NT) and PT tissue, tumor cells from liver metastatic lesions (HM) tissue exhibited significantly higher malignant phenotype\u0026nbsp;\u003csup\u003e22\u003c/sup\u003e, which was consistent with our result that the proportion of malignant epithelial cells were higher in LM samples. Compared to non-malignant epithelial cells, malignant epithelial cells were specifically regulated by IL-1, Periostin, and IFN-II signaling pathways. The expression of tumor-derived IL-1\u0026alpha; and IL-1\u0026beta; in pancreatic PDAC was associated with a lower survival rate for patients. Additionally, they were a significant part of the inflammatory cascade that triggers tumor-associated macrophages (TAMs) to secrete IL-1\u0026beta;\u0026nbsp;\u003csup\u003e23\u003c/sup\u003e, which was also verified in our research. In this study, the macrophages regulated malignant epithelial cells through the IL-1 pathway. POSTN was found to be strongly expressed in stromal cells adjacent to the pancreatic epithelial cells\u0026nbsp;\u003csup\u003e24\u003c/sup\u003e. Kanno et al. elucidated that POSTN could inhibit growth of PDAC cell\u0026nbsp;\u003csup\u003e24\u003c/sup\u003e. In contrast, some researches considered that POSTN promoted the proliferation and invasiveness of pancreatic cancer\u0026nbsp;\u003csup\u003e25, 26\u003c/sup\u003e. Accordingly, there was a great deal of disagreement regarding POSTN\u0026apos;s influence on PDAC cell growth. Further research showed that fibroblasts regulated malignant epithelial cells through the Periostin pathway (POSTN-ITGAV/ITGB5). It has been demonstrated that POSTN can regulate the proliferation of haematopoietic stem cells (HSCs) via interaction with ITGAV, moreover, the interaction between POSTN and ITGAV suppresses the FAK/PI3K/AKT pathway in HSCs, which raises p27Kip1 expression and improves the maintenance of quiescent HSCs\u0026nbsp;\u003csup\u003e16\u003c/sup\u003e. In the present study, POSTN was highly expressed in fibroblast enriched areas, ITGAV was highly expressed in epithelial cell aggregation areas, and the spatial distance between POSTN and ITGAV was close. Therefore, we hypothesized that POSTN might combine with ITGAV to affect the progression of pancreatic cancer via regulating the PI3K signaling pathway, which warrants further exploration in the future studies. Subsequently, we constructed a metastatic risk model (LM-index) using 34 DEGs enriched in the PI3K pathway between malignant and non- malignant epithelial cells. The pancreatic cancer patients with high LM-index had a poor prognosis, indicating this metastatic risk model might effectively predict the prognosis of pancreatic cancer patients.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eITGAV is an av protein-encoding gene, the product of ITGAV belongs to the integrin alpha chain family\u0026nbsp;\u003csup\u003e27\u003c/sup\u003e. It has been reported that ITGAV involved in the extracellular matrix (ECM)-receptor interaction pathway\u0026nbsp;\u003csup\u003e28\u003c/sup\u003e. The ITGAV was highly expressed in multiple cancers, such as colorectal cancer\u0026nbsp;\u003csup\u003e29\u003c/sup\u003e, gastric cancer\u0026nbsp;\u003csup\u003e30\u003c/sup\u003e and esophageal adenocarcinoma\u0026nbsp;\u003csup\u003e31\u003c/sup\u003e. The overexpression of ITGAV might be linked to the spread of breast cancer via upregulating PXN\u0026nbsp;\u003csup\u003e32\u003c/sup\u003e. Moreover, ITGAV was differentially expressed among three immune cell infiltration subtypes of metastatic pancreatic neuroendocrine tumors\u0026nbsp;\u003csup\u003e33\u003c/sup\u003e. In pancreatic cancer, ITGAV could affect the prognosis and clinicopathological factors of tumor metastasis, and the patients with high ITGAV expression exhibited inferior prognosis and recurrence rate compared to patients with low ITGAV expression\u0026nbsp;\u003csup\u003e34\u003c/sup\u003e. In addition, by upregulating the expression of ITGAV, A1B1 can promote ECM signal transduction, thereby accelerating progression of PDAC\u0026nbsp;\u003csup\u003e35\u003c/sup\u003e. Correspondingly, in PDAC cells, ITGAV knockdown significantly decreased peritoneal carcinomatosis, spontaneous lung metastasis, and primary tumor development\u0026nbsp;\u003csup\u003e36\u003c/sup\u003e. In our study, we also found that the knockdown of ITGAV significantly inhibited the invasion and migration of PANC-1 cells, indicating ITGAV under-expression might inhibit progression of pancreatic cancer cells. However, the mechanism of ITGAV affecting the progression of pancreatic cancer needs to be further explored in future studies.\u003c/p\u003e\n\u003cp\u003eIn conclusion, this study clarified possible cellular origins and drivers of LM pancreatic cancer based on the single-cell transcriptomic profiles, and discovered that ITGAV mRNA and protein were expressed in pancreatic cancer tissue and cells, and demonstrated that knockdown of ITGAV inhibited the progression of pancreatic cancer. Our results provide more reference information for understanding the mechanism of ITGAV in LM pancreatic cancer patients.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eLM\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Liver metastasis\u003c/p\u003e\n\u003cp\u003ePT\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Primary tumor\u003c/p\u003e\n\u003cp\u003eCNV\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Copy number variation\u003c/p\u003e\n\u003cp\u003eTME\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Tumor microenvironment\u003c/p\u003e\n\u003cp\u003ePDAC\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Pancreatic ductal adenocarcinoma\u003c/p\u003e\n\u003cp\u003eEMT\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Epithelial-mesenchymal transition\u003c/p\u003e\n\u003cp\u003eDEGs\u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Differentially expressed genes\u003c/p\u003e\n\u003cp\u003eGO\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Gene Ontology\u003c/p\u003e\n\u003cp\u003eKEGG\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Kyoto Encyclopedia of Genes and Genome\u003c/p\u003e\n\u003cp\u003eGSVA\u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Gene set variation analysis\u003c/p\u003e\n\u003cp\u003eBEAM\u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Branch expression analysis modeling\u003c/p\u003e\n\u003cp\u003eOCLR\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;One class logistic regression\u003c/p\u003e\n\u003cp\u003eFBS\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Fetal bovine serum\u003c/p\u003e\n\u003cp\u003esiRNA\u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Small interfering RNA\u003c/p\u003e\n\u003cp\u003eNC\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Negative control\u003c/p\u003e\n\u003cp\u003eHSCs \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; Haematopoietic stem cells\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study did not receive any specific funding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material .All data are available from the TCGA database (https://xenabrowser.net/datapages/) and GEO (https://ncbi.nlm.nih.gov/geo/) database within the article. GEO database under accession number GSE154778, GSE154778andGSE202740.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eApproval of the research protocol by an Institutional Review Board:\u0026nbsp;The use of the human tissue microarrays was approved by the Ethics Committee of Union Hospital of Tongji Medical College of Huazhong University of Science and Technology.\u003c/p\u003e\n\u003cp\u003eInformed Consent: N/A.\u003c/p\u003e\n\u003cp\u003eRegistry and the Registration No. of the study/trial: N/A\u003c/p\u003e\n\u003cp\u003eAnimal Studies: N/A.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJC and TP designed and supervised the entire work. JC, YG, TC, JX and ZK carried out the bioinformatics analysis. YL, XL and TY performed the experiments. JC, YG and JZ analysed the data and prepared the figures. JC and TP prepared and wrote the paper. All authors contributed to the article and approved the submitted version.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSiegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. \u003cem\u003eCA: a cancer journal for clinicians\u003c/em\u003e. 2022; 72: 7-33.\u003c/li\u003e\n\u003cli\u003eRitzkowsky A. [Accidental needle injuries of medical personnel]. \u003cem\u003eDtsch Med Wochenschr\u003c/em\u003e. 1994; 119: 1563-1564.\u003c/li\u003e\n\u003cli\u003eXia C, Dong X, Li H, et al. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. \u003cem\u003eChin Med J (Engl)\u003c/em\u003e. 2022; 135: 584-590.\u003c/li\u003e\n\u003cli\u003eMoore A, Donahue T. Pancreatic Cancer. \u003cem\u003eJama\u003c/em\u003e. 2019; 322: 1426.\u003c/li\u003e\n\u003cli\u003eKolbeinsson HM, Chandana S, Wright GP, Chung M. Pancreatic Cancer: A Review of Current Treatment and Novel Therapies. \u003cem\u003eJ Invest Surg\u003c/em\u003e. 2023; 36: 2129884.\u003c/li\u003e\n\u003cli\u003eTorphy RJ, Fujiwara Y, Schulick RD. Pancreatic cancer treatment: better, but a long way to go. \u003cem\u003eSurg Today\u003c/em\u003e. 2020; 50: 1117-1125.\u003c/li\u003e\n\u003cli\u003eMizrahi JD, Surana R, Valle JW, Shroff RT. Pancreatic cancer. \u003cem\u003eLancet (London, England)\u003c/em\u003e. 2020; 395: 2008-2020.\u003c/li\u003e\n\u003cli\u003eLi CG, Zhou ZP, Jia YZ, Tan XL, Song YY. Radioactive (125)I seed implantation for pancreatic cancer with unexpected liver metastasis: A preliminary experience with 26 patients. \u003cem\u003eWorld J Clin Cases\u003c/em\u003e. 2021; 9: 792-800.\u003c/li\u003e\n\u003cli\u003eDucreux M, Cuhna AS, Caramella C, et al. Cancer of the pancreas: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. \u003cem\u003eAnnals of oncology : official journal of the European Society for Medical Oncology\u003c/em\u003e. 2015; 26 Suppl 5: v56-68.\u003c/li\u003e\n\u003cli\u003e Gai B, Zhang F. Chinese expert consensus on radioactive (125)I seeds interstitial implantation brachytherapy for pancreatic cancer. \u003cem\u003eJournal of cancer research and therapeutics\u003c/em\u003e. 2018; 14: 1455-1462.\u003c/li\u003e\n\u003cli\u003e Chen K, Wang Q, Li M, et al. Single-cell RNA-seq reveals dynamic change in tumor microenvironment during pancreatic ductal adenocarcinoma malignant progression. \u003cem\u003eEBioMedicine\u003c/em\u003e. 2021; 66: 103315.\u003c/li\u003e\n\u003cli\u003e Sang W, Zhou Y, Chen H, et al. Receptor-interacting protein kinase 2 is an immunotherapy target in pancreatic cancer. \u003cem\u003eCancer discovery\u003c/em\u003e. 2023.\u003c/li\u003e\n\u003cli\u003e Zhao L, Zhao H, Yan H. Gene expression profiling of 1200 pancreatic ductal adenocarcinoma reveals novel subtypes. \u003cem\u003eBMC cancer\u003c/em\u003e. 2018; 18: 603.\u003c/li\u003e\n\u003cli\u003e Peng J, Sun BF, Chen CY, et al. Single-cell RNA-seq highlights intra-tumoral heterogeneity and malignant progression in pancreatic ductal adenocarcinoma. \u003cem\u003eCell research\u003c/em\u003e. 2019; 29: 725-738.\u003c/li\u003e\n\u003cli\u003e Ligorio M, Sil S, Malagon-Lopez J, et al. Stromal Microenvironment Shapes the Intratumoral Architecture of Pancreatic Cancer. \u003cem\u003eCell\u003c/em\u003e. 2019; 178: 160-175 e127.\u003c/li\u003e\n\u003cli\u003e Khurana S, Schouteden S, Manesia JK, et al. Outside-in integrin signalling regulates haematopoietic stem cell function via Periostin-Itgav axis. \u003cem\u003eNature communications\u003c/em\u003e. 2016; 7: 13500.\u003c/li\u003e\n\u003cli\u003e Chen W, Zheng R, Baade PD, et al. Cancer statistics in China, 2015. \u003cem\u003eCA: a cancer journal for clinicians\u003c/em\u003e. 2016; 66: 115-132.\u003c/li\u003e\n\u003cli\u003e Kim J. Cell Dissemination in Pancreatic Cancer. \u003cem\u003eCells\u003c/em\u003e. 2022; 11.\u003c/li\u003e\n\u003cli\u003e Karunakaran M, Barreto SG. Surgery for pancreatic cancer: current controversies and challenges. \u003cem\u003eFuture Oncol\u003c/em\u003e. 2021; 17: 5135-5162.\u003c/li\u003e\n\u003cli\u003e He H, Zhang S, Yang H, et al. Identification of Genes Associated with Liver Metastasis in Pancreatic Cancer Reveals PCSK6 as a Crucial Mediator. \u003cem\u003eCancers\u003c/em\u003e. 2022; 15.\u003c/li\u003e\n\u003cli\u003e Nagai M, Wright MJ, Ding D, et al. Oncologic resection of pancreatic cancer with isolated liver metastasis: Favorable outcomes in select patients. \u003cem\u003eJ Hepatobiliary Pancreat Sci\u003c/em\u003e. 2023; 30: 1025-1035.\u003c/li\u003e\n\u003cli\u003e Zhang S, Fang W, Zhou S, et al. Single cell transcriptomic analyses implicate an immunosuppressive tumor microenvironment in pancreatic cancer liver metastasis. \u003cem\u003eNature communications\u003c/em\u003e. 2023; 14: 5123.\u003c/li\u003e\n\u003cli\u003e Brunetto E, De Monte L, Balzano G, et al. The IL-1/IL-1 receptor axis and tumor cell released inflammasome adaptor ASC are key regulators of TSLP secretion by cancer associated fibroblasts in pancreatic cancer. \u003cem\u003eJ Immunother Cancer\u003c/em\u003e. 2019; 7: 45.\u003c/li\u003e\n\u003cli\u003e Kanno A, Satoh K, Masamune A, et al. Periostin, secreted from stromal cells, has biphasic effect on cell migration and correlates with the epithelial to mesenchymal transition of human pancreatic cancer cells. \u003cem\u003eInternational journal of cancer\u003c/em\u003e. 2008; 122: 2707-2718.\u003c/li\u003e\n\u003cli\u003e Erkan M, Kleeff J, Gorbachevski A, et al. Periostin creates a tumor-supportive microenvironment in the pancreas by sustaining fibrogenic stellate cell activity. \u003cem\u003eGastroenterology\u003c/em\u003e. 2007; 132: 1447-1464.\u003c/li\u003e\n\u003cli\u003e Ben QW, Jin XL, Liu J, Cai X, Yuan F, Yuan YZ. Periostin, a matrix specific protein, is associated with proliferation and invasion of pancreatic cancer. \u003cem\u003eOncol Rep\u003c/em\u003e. 2011; 25: 709-716.\u003c/li\u003e\n\u003cli\u003e Chen ZZ, Wang WP, Xue HM, Liang Y. The lncRNA-miRNA-integrin alpha V ceRNA network can affect the occurrence and prognosis of gastric cancer. \u003cem\u003eInt J Clin Exp Pathol\u003c/em\u003e. 2022; 15: 388-402.\u003c/li\u003e\n\u003cli\u003e He B, He Y, Shi W, et al. Bioinformatics analysis of gene expression alterations in microRNA‑122 knockout mice with hepatocellular carcinoma. \u003cem\u003eMol Med Rep\u003c/em\u003e. 2017; 15: 3681-3689.\u003c/li\u003e\n\u003cli\u003e Waisberg J, De Souza Viana L, Affonso Junior RJ, et al. Overexpression of the ITGAV gene is associated with progression and spread of colorectal cancer. \u003cem\u003eAnticancer research\u003c/em\u003e. 2014; 34: 5599-5607.\u003c/li\u003e\n\u003cli\u003e Zhang W, Chen Y, Qiao Z, Liu Y. Overexpression of Integrin alpha V (ITGAV) in gastric cancer and its prognostic significance. \u003cem\u003eAsian J Surg\u003c/em\u003e. 2023; 46: 5863-5864.\u003c/li\u003e\n\u003cli\u003e Loeser H, Scholz M, Fuchs H, et al. Integrin alpha V (ITGAV) expression in esophageal adenocarcinoma is associated with shortened overall-survival. \u003cem\u003eScientific reports\u003c/em\u003e. 2020; 10: 18411.\u003c/li\u003e\n\u003cli\u003e Cheuk IW, Siu MT, Ho JC, Chen J, Shin VY, Kwong A. ITGAV targeting as a therapeutic approach for treatment of metastatic breast cancer. \u003cem\u003eAm J Cancer Res\u003c/em\u003e. 2020; 10: 211-223.\u003c/li\u003e\n\u003cli\u003e Meng D, Zhao L, Liu J, Ge C, Zhang C. Identification of the Immune Subtypes for the Prediction of Metastasis in Pancreatic Neuroendocrine Tumors. \u003cem\u003eNeuroendocrinology\u003c/em\u003e. 2023; 113: 719-735.\u003c/li\u003e\n\u003cli\u003e Iwatate Y, Yokota H, Hoshino I, et al. Machine learning with imaging features to predict the expression of ITGAV, which is a poor prognostic factor derived from transcriptome analysis in pancreatic cancer. \u003cem\u003eInt J Oncol\u003c/em\u003e. 2022; 60.\u003c/li\u003e\n\u003cli\u003e Li L, Bao J, Wang H, et al. Upregulation of amplified in breast cancer 1 contributes to pancreatic ductal adenocarcinoma progression and vulnerability to blockage of hedgehog activation. \u003cem\u003eTheranostics\u003c/em\u003e. 2021; 11: 1672-1689.\u003c/li\u003e\n\u003cli\u003e Kemper M, Schiecke A, Maar H, et al. Integrin alpha-V is an important driver in pancreatic adenocarcinoma progression. \u003cem\u003eJournal of experimental \u0026amp; clinical cancer research : CR\u003c/em\u003e. 2021; 40: 214.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"ITGAV, liver metastasis, malignant epithelial cells, pancreatic cancer, single cell sequencing","lastPublishedDoi":"10.21203/rs.3.rs-4128922/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4128922/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Pancreatic cancer is a deadliest malignancy. The most common event of tumor progression in pancreatic cancer is liver metastasis, and current treatments have not produced the desired results. However, the cellular origins and drivers of liver metastasis have yet to be defined. In this study, we identified nine cell clusters in pancreatic cancer via analyzing single-cell transcriptomic profiles. Malignant epithelial cells in liver metastatic group might come from primary tumor subclones with 27 common gene copy number variation. Cell-cell communication demonstrated that malignant epithelial cells were specifically regulated by IL-1, Periostin, and IFN-II signaling pathways and fibroblasts regulated malignant epithelial cells through the Periostin pathway (POSTN-ITGAV/ITGB5). Moreover, we found that 34 differentially expressed genes were highly enriched in PI3K signaling pathway. Next, we constructed a LM index using these 34 genes, and found that pancreatic cancer patients with high-index exhibited a poor prognosis. The immuno-fluorescence staining assay showed that the spatial distance between POSTN and ITGAV was close, and ITGAV protein was expressed in tissue sections of pancreatic cancer patients. Finally, we found that ITGAV was upregulated in PANC-1 cells, and knockdown of ITGAV inhibited the invasion and migration of PANC-1 cells. This study clarified possible cellular origins and driv-ers of LM pancreatic cancer based on the single-cell and spatial transcriptomic profiles, and dis-covered that the ITGAV mRNA and protein were expressed in pancreatic cancer tissue and cells, and the knockdown of ITGAV inhibited the progression of pancreatic cancer.","manuscriptTitle":"Single-cell and spatial transcriptomic analyses reveal the cellular origins and drivers of liver metastasis from pancreatic cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-18 17:40:26","doi":"10.21203/rs.3.rs-4128922/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2bd10d6c-86ae-441f-8edd-211d3c1417e7","owner":[],"postedDate":"April 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":30631286,"name":"Biological sciences/Genetics/Genomics/Transcriptomics"},{"id":30631287,"name":"Biological sciences/Cancer/Metastases"}],"tags":[],"updatedAt":"2024-07-22T05:56:52+00:00","versionOfRecord":[],"versionCreatedAt":"2024-04-18 17:40:26","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4128922","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4128922","identity":"rs-4128922","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}