Application of angiopoietin in treatment of pancreatic cancer

Abstract

The invention provides an application of angiopoietin (ANG) in treatment of pancreatic cancer. Research finds that a novel stem cell-like cluster cell subtype (hereinafter referred to as' stem cell-like TCs') is found, and the subtype is significantly related to poor prognosis of pancreatic ductal adenocarcinoma patients. Therefore, the invention provides a novel pancreatic ductal adenocarcinoma treatment medicine, namely angiopoietin, which has very positive significance in treatment of pancreatic ductal adenocarcinoma.

Description

Technical Field

[0001] This invention relates to the field of biotechnology, and in particular to the application of angiopoietin in the treatment of pancreatic cancer. Background Technology

[0002] Pancreatic cancer is a serious disease that severely threatens human health. With societal development, its incidence and mortality rates are gradually increasing. It is projected that by 2030, pancreatic cancer will become the second leading cause of cancer-related deaths globally. The pathological classification of pancreatic cancer is based on the origin and morphological characteristics of tumor cells, mainly divided into subtypes such as ductal adenocarcinoma, adenosquamous carcinoma, and colloid carcinoma, with ductal adenocarcinoma being the most common (accounting for 85%–90%). Pancreatic ductal adenocarcinoma (PDAC) is a highly malignant tumor, and currently, there are no effective treatments.

[0003] Tuft cells are a type of epithelial cell with apical clusters of microvilli, distributed in various tissues and involved in key functions such as immune defense, sensing, and regeneration. A rare epithelial cell type, tuft cells originate from the endoderm and are primarily located on the mucosal surface of the nasopharynx, gastrointestinal tract, and airways. Tuft cells possess both sensory and secretory functions and exhibit unique morphological, functional, and molecular characteristics. Tuft cell differentiation from stem cells is regulated by the transcription factor POU2F3, and its lineage is maintained by the transcriptional repressor GFI1B. Recent studies have also found that the POU2F3 binding factors OCA-T1 (POU2AF2) and OCA-T2 (POU2AF3) are crucial to the tuft cell lineage. Tuft cells express taste receptors, including TRPM5, α-taste receptor protein (GNAT3), and Tas1 / 2R, which help in the perception of bitter, umami, and sweet taste molecules. In the gut, cluster cells are divided into Tuft1 and Tuft2 subtypes: Tuft1 exhibits neuronal-related transcriptional programs, while Tuft2 shows immune-related programs. Their spatial distribution also differs, with Tuft2 markers enriched at the villus tips and Tuft1 markers located at the villus base.

[0004] Cluster cells (TCs) are a widely distributed but functionally heterogeneous type of epithelial cell, exhibiting significant distribution and functional diversity in organs such as the respiratory and digestive systems. Although present under homeostatic conditions in tissues like the intestine and thymus, they are almost entirely absent in the normal pancreas, and their presence is a hallmark of chronic pancreatitis, observed in approximately 27% of human cases. Regarding the role of TCs in the development of pancreatic ductal adenocarcinoma (PDAC), conflicting reports exist, with some suggesting an inhibitory effect and others a promoting one. The relationship between the two and the mechanism of action remain unclear. Summary of the Invention

[0005] In view of the above-mentioned existing views, the present invention identifies and characterizes for the first time a novel subtype of stem cell-like TCs (hereinafter referred to as "stem cell-like TCs") that plays a key pathogenic role in PDAC, expressing SUCNR1, CD133 and IL4 / 13 receptors, and this subtype is significantly associated with poor prognosis in patients.

[0006] This invention provides, in one aspect, the application of stem cell-like TCs inhibitors in the preparation of drugs for treating pancreatic ductal adenocarcinoma. The stem cell-like TCs inhibitors are formulations capable of inhibiting their proliferation and / or survival.

[0007] Furthermore, the stem cell-like TCs inhibitor is a chemical substance capable of binding to its receptor, such as a functional protein.

[0008] Furthermore, the receptor is PLXNB2 and / or CEACAM1.

[0009] Furthermore, the stem cell-like TCs inhibitor is angiopoietin (ANG) or a protein with at least 85%, 90%, or 95% of the same functionality.

[0010] Another aspect of the invention provides the use of angiopoietin (ANG) or a protein having at least 85%, 90% or 95% of the same functionality as ANG in the preparation of a medicament for treating pancreatic ductal adenocarcinoma.

[0011] Another aspect of the present invention provides the use of PLXNB2 inhibitors in the preparation of treatments for pancreatic ductal adenocarcinoma.

[0012] Furthermore, the PLXNB2 inhibitor is angiopoietin (ANG) or a protein with at least 85%, 90%, or 95% of the same functionality.

[0013] Another aspect of the present invention provides the application of a SUCNR1 detection reagent in the preparation of drugs for the diagnosis, classification, or prognostic assessment of pancreatic ductal adenocarcinoma, wherein the SUCNR1 detection reagent is a reagent for detecting the expression level of SUCNR1 in a sample.

[0014] Another aspect of the present invention provides the application of a CD133 detection reagent in the preparation of a drug for the diagnosis, classification or prognostic assessment of pancreatic ductal adenocarcinoma, wherein the CD133 detection reagent is a reagent for detecting the expression level of CD133 in a sample.

[0015] One aspect of the present invention provides the use of reagents for detecting SUCNR1 or CD133 in the preparation of drugs for the diagnosis, typing, or prognostic assessment of pancreatic ductal adenocarcinoma.

[0016] Another aspect of the present invention provides a medicament for treating pancreatic ductal adenocarcinoma, wherein the main component of the medicament is a stem cell-like TCs inhibitor. The medicament also contains excipients acceptable to the human body.

[0017] Another aspect of the present invention provides a pharmaceutical composition comprising the drug as described above.

[0018] Another aspect of the present invention provides a medicament for treating pancreatic ductal adenocarcinoma, wherein the main component of the medicament is a PLXNB2 inhibitor. The medicament also contains excipients acceptable to the human body.

[0019] Another aspect of the present invention provides a pharmaceutical composition comprising the drug as described above.

[0020] Another aspect of the present invention provides the application of PDAC patient-derived organoid models comprising the said stem cell-like TCs in screening diagnostic or therapeutic drugs for pancreatic ductal adenocarcinoma.

[0021] As described above, the angiopoietin disclosed in this invention has the following beneficial effects: Research has discovered that inhibiting a novel cluster cell subtype found in pancreatic ductal carcinoma can effectively suppress pancreatic ductal carcinoma, especially by using angiopoietin (ANG) to inhibit tumor cell proliferation and survival. Attached Figure Description

[0022] Figure 1 A comprehensive cell type annotation and clustering visualization of cell populations from scRNA sequencing data of human PDAC tissues; Figure 2 It reveals subgroups of unique transcriptional profiles within epithelial cells; Figure 3 It shows specific markers of TC; Figure 4 This demonstrates the association between TC subsets and intestinal Tuft2 cells; Figure 5 The expression levels of prostaglandin D synthase and COX1 in the TC subset were shown; Figure 6 The differentially expressed genes were shown among the TC subgroups; Figure 7 This showed an association between TC subsets and intestinal Tuft1 cells; Figure 8 The expression of stem cell-like TC subset markers was shown; Figure 9 The gene set enrichment analysis of the TC subset is shown; Figure 10This showed that Tuft2 cells were enriched in antigen presentation; Figure 11 It showed the transcriptional activity of TC markers and cell-specific transcription factors; Figure 12 This demonstrates the differences in transcription factor activity between the two TC isoforms; Figure 13 The expression of transcription factors POU2F3 and GFI1B and cofactors POU2AF2 and POU2AF3 in two TC isotypes was shown. Figure 14 The stemness gene set was shown to assess the stemness of two TC subtypes; Figure 15 The distribution scores of Tuft2 cells in various subtypes of pancreatic cancer are shown; Figure 16 This study showed the relationship between two TC subtypes and the survival of pancreatic cancer patients; Figure 17 The results of multiplex immunofluorescence staining using DCLK1 as a TC marker are shown. Figure 18 The pathological zoning of a pancreatic cancer tissue section is shown; Figure 19 The distribution statistics of different cell types in pancreatic cancer tissue sections are shown; Figure 20 A visualization analysis of the distribution characteristics of stem cell-like TCs and Tuft2 cells in pancreatic cancer tissue is shown. Figure 21 The distribution maps of stem cell-like TCs and Tuft2 in the Ductal1 and Ductal2 hotspot regions are shown. Figure 22 The colocalization relationships of stem cell-like TCs and Tuft2 with Ductal2, Ductal1 and acinar cells were shown; Figure 23 The colocalization of stem cell-like TCs with Ductal1 and Tuft2 with Ductal2 is shown in the Ductal1 hotspot region and the Ductal2 hotspot region. Figure 24 It showed a highly active ligand-receptor relationship between stem cell-like TCs and ductal cells; Figure 25 This demonstrates the evaluation of predicted ligand-receptor combination activity using the stLearn algorithm; Figure 26 Immunostaining images of stem cell-like TCs after the addition of various cytokines during organoid culture are shown. Figure 27The percentage of stem cell-like TCs was shown after the addition of various cytokines during organoid culture. Detailed Implementation

[0023] Currently known tumor cells (TCs) are epithelial cells that secrete prostaglandin D2 synthase (PGD2) and play a role in tumor suppression. In this study, we first identified a TC subpopulation with a transcriptional profile similar to the intestinal Tuft2 subtype, characterized by high expression of HPGDS, MHC class I molecules, and COX1 (named Tuft2 subtype or Tuft2 cells). Furthermore, we identified a novel TC subpopulation: a new stem cell-like TC subpopulation expressing SUCNR1, CD133, and IL4 / 13 receptors, exhibiting characteristics consistent with tumor-promoting and stem cell-like properties (hereinafter referred to as "stem cell-like TCs" or "stem cell-like TCs"). Using Visium data, we spatially localized these two cell types, revealing their discrete distribution characteristics—Tuft2 cells are primarily located in the malignant ductal epithelium (Ductal2), while stem cell-like TCs are located in the non-malignant ductal epithelium (Ductal1). In summary, this study systematically elucidates the role of TCs and provides new insights into the spatial heterogeneity of this rare but important cell type and its role in tumor progression.

[0024] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0025] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0026] Abbreviations TCs: clustered cells scRNA-Seq: Single-cell RNA sequencing IF: Immunofluorescence PDAC: Pancreatic ductal adenocarcinoma GSEA: Gene Set Enrichment Analysis MIA: Multimodal Integration Analysis ST: Spatial Transcriptomics The original expression matrix of PDAC (n=35) used for single-cell RNA sequencing was obtained from the Genome Sequence Archive (GSA), accessible at https: / / bigd.big.ac.cn / gsa, accession number CRA0011601. Visium spatial transcriptomics results of tumors from 4 PDAC patients were obtained from the Gene Expression Comprehensive Database (GEO) (GSE235315). For immunofluorescence staining, the collection of patient tumor tissue samples and the establishment of PDAC surgical tumor samples for PDOs were authorized by the Ethics Committee of Fujian Provincial Hospital (K2021-08-004, Fuzhou, China). Written informed consent was obtained from all patients.

[0027] Sc-RNA-seq data quality control, integration and dimensionality reduction We used Seurat 4.4.0(8) software to load the gene expression matrix, and then filtered out low-quality cells (<200 transcripts / cell, mitochondrial gene percentage >10%) and low-expression genes (<3 cells / gene). The original expression matrix was normalized and scaled using SCTransform to generate 3000 highly variable genes for principal component analysis (PCA) dimensionality reduction. Subsequently, we used the Harmony software package to remove batch effects between samples. Finally, we used the previously calculated top 20 principal components and applied UMAP (Uniform Manifold Approximation and Projection) for visualization of single-cell clustering.

[0028] Visium Spatial Transcriptomics Data Quality Control and Analysis Visium spatial transcriptomics data were processed using the Scanpy software package. Low-quality data (defined as data expressing fewer than 500 genes) were filtered out from subsequent downstream analyses. The filtered data were then normalized and integrated using Harmonypy to eliminate batch effects between samples. Cluster analysis was performed using the Leiden algorithm (resolution = 0.5). Based on the pathological assessment results of matched tissue sections, the resulting clusters were annotated as different tissue regions.

[0029] Gene enrichment analysis To investigate the functional differences between the two cluster cell subpopulations and the region-specific functional changes in the co-localization of cluster cell and duct cell subpopulations, we used the FindMarkers Seurat function for differentially expressed genes (DEG) analysis. We downloaded the Wikipathway and GO gene sets from the Molecular Marker Database (MSigDB) (https: / / github.com / igordot / msigdbr) as reference gene sets, accessible at http: / / software.broadinstitute.org / gsea / msigdb / . Using the clusterProfiler R package, we performed gene set enrichment analysis (GSEA) and visualized the functional characteristics of the differentially expressed genes.

[0030] TCGA Analysis Using the UCSC data portal (https: / / xenabrowser.net), we downloaded batch RNA sequencing results and survival data from the TCGA pancreatic adenocarcinoma (PAAD, n=182) dataset and retained 150 PDAC patients for further analysis. We screened for differentially expressed genes with adjusted p-values ​​<0.01 and |log2FC|>1, and used positive and negative log2 fold changes as coefficients for each gene. We calculated a score for each patient by multiplying their gene expression levels by these coefficients. Based on this score, patients were divided into two groups. We used the survminer (version 0.4.8) and survival (version 3.5.5) R packages to plot Kaplan-Meier survival curves and cumulative event count tables. To investigate the potential association between cluster cells and different molecular subtypes of pancreatic cancer, we calculated a Tuft2 score for each patient. For the analysis of PDAC subtypes, data were directly obtained from the supplementary material by Bailey et al. We then compared the distribution of Tuft2 scores across the four different molecular subtypes.

[0031] pySCENIC analysis To determine the activity of each intracellular regulator, we used pySCENIC for analysis and evaluated it using AUCell within the scenic framework. Subsequently, we used the FindMarkers function to identify transcription factors with differential activity between the two cell types using the Wilcoxon rank-sum test (adjusted p-value threshold <0.05). A heatmap was finally generated to present the results.

[0032] Spatial deconvolution using cell2location to analyze the distribution of clustered cell subtypes To analyze the spatial distribution characteristics of Tuft cell subtypes (stem cell-like TCs and Tuft2), we employed the reference genome-based spatial localization tool cell2location. Using an annotated single-cell RNA sequencing dataset as a reference, we strictly followed the authors' recommended workflow to deconvolve the Visium spatial transcriptome data. After gene loading and filtering of the Visium dataset using standard parameters, we initialized the model using the cell feature matrix generated from the single-cell RNA sequencing reference data and performed an initial training run of 250 epochs. We then extracted common genes across the datasets and set the deconvolution parameters to: cell_per_location=30 and detection_alpha=20. Finally, the cell2location model was trained for 30,000 epochs, and the 5th percentile of the posterior distribution of cell abundance estimation was exported to the annadata project for use in all subsequent downstream analyses.

[0033] Multimodal integration analysis MIA used hypergeometric tests to assess the association between each spatial transcriptomics (ST) cluster and each single-cell RNA sequencing (scRNA-seq) cluster. For each ST cluster, the 200 most significantly upregulated genes (adjusted p-value ≤ 0.05 and log2 FC > 1) were selected to infer the enriched cell types within each spatial region.

[0034] Hotspot identification and mistyR analysis First, we used the spottedpy package to identify hotspot regions for Ductal1 and Ductal2 cells. Based on the cell2location deconvolution analysis results, we further defined the top 10% of spots with the highest abundance of stem cell-like TCs and Tuft2 as their respective hotspot regions, thereby achieving co-localization analysis of ductal cells and the Tuft subpopulation. To quantitatively assess cell co-localization patterns in different microenvironments, we extracted the deconvolution cell abundance matrices of the Ductal1 and Ductal2 hotspot regions and used them as input data for mistyR. By running the run_misty function, we estimated the spatial co-localization intensity between cell types within each ductal hotspot region and further compared the differences in co-localization intensity of specific cell pairs in the two ductal regions to identify environment-dependent spatial relationships.

[0035] Immunofluorescence staining Pathological tissues and organoids (PDOs) were fixed in formalin, embedded in paraffin, and cut into 4-micrometer thick sections. After dewaxing, the sections were retrieval with EDTA antigen, followed by inactivation and blocking. Primary antibody was incubated overnight at 4°C, followed by secondary antibody incubation at room temperature for 30 minutes. Sections were stained with dye for 10 minutes, followed by heating with antigen elution buffer for 20 minutes, repeated three times. Finally, DAPI incubation and mounting were performed. All reagents except the primary antibody were purchased from the Four-Target Five-Color Fluorescence Kit (Immunoway RS0037). Primary antibodies used for immunofluorescence staining include: rabbit anti-HLA class I ABC (1:5000, Proteintech 66013-1-Ig), rabbit anti-CD133 (1:5000, Proteintech 18495-1-AP), rabbit anti-SUCNR1 (1:5000, Immunoway YT2035), rabbit anti-HPGDS (1:5000, Proteintech 22522-1-AP), rabbit anti-COX1 (1:5000, Proteintech 13393-1-AP), and rabbit anti-DCAMKL1 (1:8000, abcam ab109029).

[0036] NicheNet ligand-receptor analysis and stLearn validation To investigate intercellular communication, we performed ligand-receptor (LR) interaction analysis using NicheNet. The default threshold for NicheNet was set to log2FC = 0.15, with adjusted p-values ​​≤ 0.05. Ligand-receptor pairs were ranked according to priority scores, and the 30 most active pairs were selected for further spatial validation. The activities of these ligand-receptor pairs were calculated using stLearn with the st.tl.cci.run() function. To minimize potential interference from Tuft2 cells, we identified spatial regions with low Tuft2 abundance and high stem cell-like TC abundance based on spatial deconvolution results. LR interaction activity was visualized and analyzed within these specific regions.

[0037] PDAC PDO culture 1) Clean the tumor tissue, mince it, and then digest it with a mild cell dissociation reagent (GCDR) (GCDR; #7174, stemcell) at 37°C for 45-60 minutes to obtain single cells. Then aspirate the supernatant and neutralize it.

[0038] 2) The cells were collected by centrifugation (500g, 5min), counted, and 2000 cells were mixed with matrix gel and seeded in 24-well plates.

[0039] 3) Mix the components of PancreaCult™ organoid formation initiation medium (OIM; #100-0820, stemcell), supplemented with PGE2 (PGE2; #72192, stemcell) and Y-27632 (#72302, stemcell); three days later, replace the initiation medium with PancreaCult™ organoid formation growth medium (OGM; #100-0781, stemcell). Expand and cryopreserve the organoids according to the CS10 protocol (#7931, stemcell) for subsequent research.

[0040] statistics Cox regression analysis was used to assess the association between TC subtypes and overall survival. Wilcoxon tests were used to identify biomarkers for TC subtypes, and differential gene expression and transcription factor activity among TC subgroups were compared. The abundance of TC subgroups in different ductal regions was assessed, and the distribution of TC subgroups in pancreatic cancer subtypes was analyzed. First, the Kruskal-Wallis test was used to assess overall differences among the four subtypes, followed by pairwise comparisons using the Wilcoxon test. Statistical significance was defined as P < 0.05.

[0041] Example 1: Discovery and characterization of two new subpopulations of pancreatic cancer cluster cells We used Seurat 4.4.0 software to load the gene expression matrix, then filtered out low-quality cells (<200 transcripts / cell, mitochondrial gene percentage >10%) and low-expression genes (<3 cells / gene). The original expression matrix was normalized and scaled using SCTransform to generate 3000 highly variable genes for principal component analysis (PCA) dimensionality reduction. Subsequently, we used the Harmony software package to remove batch effects between samples. Finally, we used the previously calculated top 20 principal components and applied UMAP (Uniform Manifold Approximation and Projection) for single-cell cluster visualization.

[0042] Comprehensive cell type annotation and clustering of scRNA sequencing data from human PDAC tissues revealed the expected cell populations, including fibroblasts, T cells, B cells, macrophages, endothelial cells, and endocrine cells, confirmed by the expression of cell marker genes. As previously studied, ductal epithelial cells were classified into two main types: Ductal Cell Type 1 and Ductal Cell Type 2. Figure 1 Within the epithelial cells, further subset analysis identified the TC cell population (). Figure 2The identification of TC cell populations was based on the distribution of specific markers POU2F3, GNAT3, TRPM5, PTGS1, and DCLK1. Figure 3 Overall, PDAC TCs are rare, accounting for 2.32% of the total epithelial cell population (range: 0.21%–20.95%).

[0043] To characterize the new TC subsets, we used existing intestinal epithelial TC cell classifications. One subset of pancreatic TCs exhibited a transcriptional profile consistent with intestinal Tuft2 characteristics. Figure 4 It is characterized by the expression of HPGDS, COX1 (PTGS1), ALOX5, CD300LF, IL17RB, ALOX5AP, and MHC class I / II subunits (HLA-A / B / C, HLA-DR / QB1). Figure 5 , 6). However, the second subgroup showed a lower association with the previously described Tuft1 gene signature ( Figure 7 This suggests that the group may represent a distinct subtype of pancreatic cancer TC. Overall, Tuft2 cells comprised an average of 0.85% of the epithelium (range: 0–4.42%), while the novel TC subtype comprised an average of 1.18% of the epithelium (range: 0–20%).

[0044] Example 2: Molecular differences and identification between two cluster cell subtypes Based on the analysis results of Example 1, we performed differential gene expression analysis based on cell type and further performed the same analysis between the two Tuft cell subpopulations. We identified SUCNR1, CD133 (PROM1), and SOX4 as key biomarkers for this new subgroup. Figure 8 、 and 6).

[0045] Notably, CD133 and SOX4 are associated with pancreatic cancer stem cells and tumor initiation, with SOX4 being crucial for the expansion of DCLK1+ TCs in precancerous lesions. This suggests that this subtype may possess a stem cell-like phenotype, which we term "stem cell-like TCs." In contrast, HPGDS are associated with tumor suppressor function and are primarily expressed in Tuft2 cells. Gene set enrichment analysis (GSEA) revealed that Tuft2 cells were enriched in antigen presentation and prostaglandin signaling, indicating that these functions are specific rather than broadly applicable to cluster cells as previously thought. The function of stem cell-like TCs is strongly associated with pancreatic cancer subtypes, EMT, and metabolic reprogramming in pancreatic cancer, supporting their pro-tumorigenic function. Figure 9-10 ).

[0046] Example 3: Study on the functional activity of two TC subtypes of cells We used pySCENIC for analysis and assessed the activity of each transcription factor using AUCell. Subsequently, we used FindMarkers to identify differentially expressed regulators between the two Tuft cell types using the Wilcoxon rank-sum test (adjusted p-value threshold <0.05). We confirmed that the major TC regulators POU2F3 and GFI1B exhibited strong activity in both TC subtypes, highlighting the accuracy of this analytical technique. Figure 11 POU2AF1 is a transcriptional cofactor of POU2F1 / 2 and a homolog of POU2AF2 / 3, a transcriptional cofactor that activates POU2F3. It exhibits enrichment of activity along with POU2F3 and POU2F2 in Tuft2 cells. Higher transcriptional activity of SOX4 and the pluripotency factor OCT4 (POU5F1) in stem cell-like TCs further supports the stem cell-like phenotype. Finally, POU2AF3 is expressed at higher levels in stem cell-like TCs. Figure 12 (13) emphasized that differences in transcription factor activity may affect the differentiation of TC subtypes.

[0047] Based on the previous description, we used the AddModuleScore method to assess the stemness of the two TC subtypes, using a 27-gene set from StemChecker (http: / / stemchecker.sysbiolab.eu / ). Stem cell-like TCs showed higher stemness scores, while Tuft2 showed the opposite in most cases. Figure 14 These results indicate that stem cell-like TC is a novel TC subtype with stem cell-like characteristics.

[0048] Example 4: Distribution of TCs in PDAC subtypes and their impact on prognosis To further investigate the role of TCs in PDAC subtypes, we used gene signatures derived from the TC subgroups obtained in Example 2. We used the positive and negative log2 fold changes of these differentially expressed genes as coefficients for each gene. A score was calculated for each patient by multiplying their gene expression levels by these coefficients and applied to 150 PDAC patients in the TCGA classified according to the subtypes of Bailey et al. The Tuft2 signature was significantly increased in immunogenic PDAC and decreased in the more aggressive non-immunogenic subtypes. Notably, squamous PDAC, the type with the worst prognosis, had the lowest Tuft2 signature. Figure 15 ).

[0049] To assess the clinical significance of the two TC subtypes, we evaluated the prognostic impact of TC subtypes on patients in the TCGA pancreatic cancer (PAAD) cohort based on the aforementioned scores. Patients with higher Tuft2 scores had significantly longer overall survival compared to those with higher stem cell-like TC scores (P=0.0084, HR=0.5651, 95% CI [0.370-0.864]). Figure 16 This further emphasizes the adverse prognostic effect of stem cell-like TCs on PDAC patients.

[0050] Example 5: Reconfirmation of Two TC Subtypes To confirm our findings, we performed multiplex immunofluorescence staining. Using DCLK1 as a pan-cluster cell marker, we identified a distinct ductal subset exhibiting SUCNR1 / CD133 / DCLK1 co-staining characteristics. Notably, these cells showed low or undetectable COX1 expression, consistent with our previous findings. Figure 17 A) supports the existence of stem cell-like TCs as a novel subtype of pancreatic TCs. Consistent with our GSEA and expression profiling results, Tuft2 cells showed co-localization of MHC class I (HLA-A / B / C), HPGDS, and COX1, suggesting their immune-related functions. Figure 17 B).

[0051] Example 6 Spatial distribution of two TC subtypes in PDAC We used PDAC spatial transcriptome (ST) data from the 10X Visium platform and excluded low-quality data (gene count <500). After removing batch effects using HarmonyPy, we performed clustering and dimensionality reduction analyses. Based on the histological structure of pancreatic cancer, the tissue sections were successfully divided into six regions: ducts, immune infiltration zone, stroma, blood vessels, islet cell proliferation zone, and leiomyoma zone. Figure 18 To validate the cell type distribution within these six regions, we applied the MIA method, selecting the 200 most significant genes in each region (adjusted p-value ≤ 0.05 and log2 FC > 1) and inferring the cell type distribution within each region. MIA results showed the presence of T cells, macrophages, cluster cells, and ductal type 2 cells in the immune infiltration zone; ductal type 1, ductal type 2, and cluster cells in the ductal zone; and macrophages, endothelial cells, and endocrine cells in the islet cell proliferation zone. Figure 19 ).

[0052] To further investigate the distribution characteristics of stem cell-like clusters and Tuft2 cells in pancreatic cancer tissue, we used Cell2Location technology to perform deconvolution analysis on each point based on annotated single-cell RNA sequencing data, with deconvolution parameters set to cell_per_location=30 and detection_alpha=20. To clarify the spatial localization of the two ductal cell types and two Tuft cell subpopulations, we analyzed and visualized the spatial distribution of duct type 1, duct type 2, stem cell-like clusters, and Tuft2 cells. The results showed that although stem cell-like TCs and Tuft2 cells were mainly located in the ductal region (… Figure 20 (above), but the two types exhibit a highly mutually exclusive distribution pattern in most ducts. Further observation revealed that some ducts are almost entirely composed of duct type 1 or duct type 2 cells, while other ducts show a mixed distribution (above). Figure 20 (See below). Specifically, stem cell-like TCs were mainly enriched in areas dominated by type 1 ductal cells, but rarely appeared in areas enriched by type 2 ductal cells. Conversely, Tuft2 cells were mainly distributed in areas enriched by type 2 ductal cells, but were almost undetectable in areas dominated by type 1 ductal cells. In areas where type 1 and type 2 ductal cells coexisted, both cell types were found simultaneously. Figure 21 ).

[0053] We used SpottedPy software to identify Ductal1 and Ductal2 hotspot regions. Given the sparse distribution of Tuft cells, we defined hotspots for stem cell-like Tuft and Tuft2 as the top 10% of cell abundance. After successful identification of Ductal1 and Ductal2 hotspots, Ductal1 hotspots overlapped with stem cell-like Tuft hotspots, and Ductal2 hotspots overlapped with Tuft2 hotspots. Furthermore, stem cell-like Tuft and Tuft2 cells exhibited spatial repulsion. Figure 20 This exclusion phenomenon is related to the ratio of Ductal1 and Ductal2 in each cell site. To further investigate the colocalization relationship between Ductal1 and stem cell-like Tuft, and between Ductal2 and Tuft2, and its correlation with the abundance of Ductal1 and Ductal2 in each site, we extracted Ductal1 hotspot and Ductal2 hotspot regions and assessed the spatial colocalization intensity between cell types within each region using mistyR. In the Ductal1 hotspot region, we observed that Ductal1 cells could predict the abundance of stem cell-like TC, while Ductal2 cells could predict the abundance of Tuft2 cells. Furthermore, stem cell-like TC could also predict the abundance of Ductal2, Ductal1, and acinar cells, while Tuft2 only colocalized with Ductal2 cells. Figure 22Similar colocalization patterns were also observed in Ductal2 hotspot regions. Notably, spatial dependence between stem cell-like TCs and Ductal1 cells was enhanced in regions with high Ductal2 cell abundance; conversely, dependence between Ductal2 and Tuft2 cells was weakened. Figure 23 ).

[0054] In summary, stem cell-type TCs exhibit spatial dependence with Ductal1 cells, as do Tuft2 and Ductal2 cells. This dependence is regulated by the relative abundance of Ductal1 and Ductal2 cells within each foci.

[0055] Example 7: IL-25, angiopoietin, and CEACAM5 can reduce stem cells in organoids derived from PDAC patients. Cellular TCs Although our spatial transcriptome analysis showed that stem cell-like TCs were mainly located in non-malignant ducts rather than malignant ducts, Visium data showed that the co-localization relationship between Ductal2 cells and stem cell-like TCs weakened with increasing Ductal2 cell abundance. Figure 22 Based on these observations, we hypothesized that Ductal2 cells might inhibit the abundance of stem cell-like TCs. Given that both clustered cell subpopulations are distributed in the ductal region, we extracted Ductal1 cells, Ductal2 cells, and stem cell-like TCs from single-cell RNA sequencing data and used NicheNet software to investigate potential ligand-receptor (LR) interactions between Ductal2 cells and stem cell-like TCs. The default threshold for NicheNet was set to log2FC = 0.15, and the adjusted p-value was ≤ 0.05. We also screened for highly active ligand-receptor combinations (…). Figure 24 (Left figure) Visual analysis of the expression levels of these highly active ligands in two types of ductal cells ( Figure 24 (middle image), and further show the receptors with the highest interaction potential with the top 30 ligands ( Figure 24 (Right figure). To reduce the potential impact of Tuft2 cells on stem cell-like TCs, we subsequently selected ductal regions with low Tuft2 abundance but high stem cell-like TC abundance. In these regions, we used the stLearn algorithm to assess the predicted ligand-receptor combination activity, aiming to precisely screen for potential ligand-receptor combinations that Ductal2 cells may regulate stem cell-like TCs. Using this method, we identified the ANG-PLXNB2 and CEACAM5-CEACAM1 ligand-receptor combinations as exhibiting high activity in these regions (…). Figure 25 ).

[0056] To investigate the role of these ligands, we used organoids (PDOs) derived from human PDAC patients and confirmed the presence of stem cell-like TCs in this model by immunofluorescence staining for CD133, SUCNR1, and DCLK1. Figure 25 IL4 / IL13 stimulation significantly increased the abundance of stem cell-like TCs. Compared with basal medium, IL4 / 13-supplemented medium increased the proportion of stem cell-like TCs by 383% (p=0.0093). Furthermore, in EGF-deficient medium supplemented with IL4 / 13, the proportion of stem cell-like TCs increased by 2.53 times compared with medium containing both EGF and IL4 / 13 (p<0.0001). This finding confirms that EGF may have an inhibitory effect on the expansion of stem cell-like TCs. Compared with the control group, the addition of cluster cell regulators IL25, angiopoietin (ANG), or CEACAM5 significantly inhibited the abundance of stem cell-like TCs. Specifically, compared with the control group, IL25 supplementation significantly reduced the abundance of stem cell-like TCs by 70.1% (p<0.0001). Supplementation with ANG reduced the abundance of stem cell-like TCs by 36.7% (p=0.0017), while supplementation with CEACAM5 led to a 34.7% reduction (p=0.0032). IL25 showed the most significant inhibitory effect on stem cell-like TCs. Figure 26 and 27 ) in conclusion In this study, we demonstrated for the first time the existence of distinct TC subsets within PDAC, including an immunomodulatory, tumor-suppressive subtype (Tuft2; expressing HPGDS, COX1, and MHC class I molecules) and a stem cell-like TC subtype (expressing SOX4, CD133, SUCNR1 receptors, and IL4 / 13 receptors). Visium analysis revealed that Tuft2 and stem cell-like TCs are spatially discrete, co-localizing with malignant and non-malignant ductal cells, respectively. Finally, we demonstrated that IL25, ANG, and CEACAM5 inhibit the proliferation of stem cell-like TCs. In summary, our findings suggest that previously reported functional differences in pancreatic cluster cells are due to spatial and clinical contextual variations in these discrete subtypes. These findings reconcile previously contradictory research.

[0057] The above embodiments are for illustrating the implementation schemes disclosed in this invention and should not be construed as limiting the invention. Furthermore, various modifications listed herein, as well as variations in the methods and compositions of the invention, will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been specifically described in conjunction with various specific preferred embodiments, it should be understood that the invention should not be limited to these specific embodiments. In fact, various modifications as described above that are obvious to those skilled in the art to obtain the invention should be included within the scope of this invention.

Claims

1. Use of a stem cell-like TCs inhibitor in the preparation of a medicament for treating pancreatic ductal adenocarcinoma.

2. Use according to claim 1, characterized in that: The stem cell-like TCs inhibitor is an agent capable of inhibiting its proliferation and / or survival.

3. Use according to claim 2, characterized in that: The stem cell-like TCs inhibitor is a chemical capable of binding to its receptor.

4. The use according to claim 1, characterized in that: The stem cell-like TCs inhibitor is angiopoietin or a functional protein having at least 85%, 90% or 95% identity thereto.

5. Use of a PLXNB2 inhibitor in the preparation of a medicament for treating pancreatic ductal adenocarcinoma.

6. Use according to claim 5, characterized in that: The PLXNB2 inhibitor is angiopoietin or a functional protein having at least 85%, 90% or 95% identity thereto.

7. A medicament for treating pancreatic ductal adenocarcinoma, characterized by, The main active ingredient of the medicament is a stem cell-like TCs inhibitor.

8. A medicament for treating pancreatic ductal adenocarcinoma, characterized by, The medicament is a medicament capable of binding to PLXNB2 and inhibiting its expression.

9. A pharmaceutical composition, characterized by, The pharmaceutical composition contains the medicament as claimed in claim 7.

10. A pharmaceutical composition, characterized by, The pharmaceutical composition contains the medicament as claimed in claim 8.

Images