Prognostic marker and prognostic prediction model for EBV-related malignant tumors and application

Abstract

The invention belongs to the technical field of biology, and particularly relates to a prognosis marker and a prognosis prediction model for EBV-related malignant tumors and application of the prognosis marker and the prognosis prediction model. According to the invention, GSE51575 transcriptome sequencing data are collected, differential expression genes are screened, and function enrichment analysis is carried out, so that a gene set of gastric cancer related to EBV infection is determined. A prognosis model based on 25 repeated DEG is constructed, and immune correlation is evaluated. The research of the invention proves that ENTPD8 and PCOLCE2 are a target spot and a biomarker of EBVaGC immunotherapy; in addition, bleomycin inhibits the development of EVBaGC by regulating and controlling the expression of ENTPD8 and PCOLCE2. The invention provides an important viewpoint for clinical research of treating EBVaGC by combining immunotherapy with BLM in future.

Description

Technical Field

[0001] The invention belongs to the field of biotechnology, and specifically relates to a prognostic marker, a prognostic prediction model and an application of an EBV-related malignant tumor. Background Art

[0002] Gastric cancer (GCs) is a major concern worldwide and has become the second leading cause of cancer-related death, with 95,000 newly diagnosed patients each year, an overall 5-year survival rate of 20%, and a median survival of less than 12 months. Approximately 9% of GCs are identified as Epstein-Barr virus-positive. Epstein-Barr virus (EBV), also known as human herpesvirus 4 (HHV4), is a double-stranded DNA virus that infects more than 90% of adults and was first discovered in Burkitt's lymphoma. Multiple factors driven by EBV contribute to tumorigenesis, including hypermethylation of tumor suppressor genes, alterations in mucosal inflammation, host immune evasion, and dysregulation of cell cycle pathways.

[0003] EBV-associated gastric cancer (EBVaGC) may be caused by the monoclonal proliferation of EBV-infected cells, which are mainly confined to the upper stomach and are often accompanied by mucosal atrophy and intestinal metaplasia. Existing studies have shown that EBVaGCs have a significantly lower frequency of lymph node metastasis (LNM) and a lower mortality rate compared with EBV-negative controls (EBVnGCs). In addition, the molecular characteristics of EBVnGCs also include focal amplification of 9p24.1 (including JAK2, PD-L1, and PD-L2), recurrent PIK3CA mutations, extreme DNA hypermethylation (CDKN2A, p16INK4A), and increased immune infiltration. This suggests that EBVaGC may escape immune recognition of T lymphocytes through the PD-1 / PD-L1 pathway, and it may be considered the root cause of the response of EBVaGC patients to immunotherapy (immune checkpoint blockade, ICB).

[0004] Patients with advanced gastric cancer are often treated with moderate chemotherapy to slow tumor progression. Although known patient data are limited, preclinical data show that EBVaGC is resistant to current chemotherapy regimens, including docetaxel or 5-FU. EBVaGC is a special subtype with abundant PD-L1 / PD-L2 expression. EBV infection induces immune responses, recruits immune cells, and regulates the expression of immune-related molecules. Compared with EBVnGC, EBVaGC has a higher proportion of lymphoepithelioma-like tumors and a better survival rate. Dysregulated immune response genes in EBVaGC show increased levels of lymphocyte infiltration and increased activation of the immune checkpoint pathway (PD-1 / PD-L1). In EBVaGC patients, the PD-L1+ group has a "hot" immune microenvironment, less aggressive clinicopathological features, better prognosis, and higher sensitivity to anti-PD-1 immunotherapy such as pembrolizumab, nivolumab, and avelumab.

[0005] Immunotherapy mainly induces antitumor effects by regulating the tumor microenvironment (TME), which usually relies on the dynamic interaction between tumor cells and immunomodulators. TME is a complex ecosystem composed of multiple cell types (tumor infiltrating lymphocytes) and other non-cellular components from the extracellular matrix. It has obvious heterogeneity and plays a key role in tumorigenesis and immunotherapy effects. However, the current dilemma faced by immunotherapy is tolerance caused by different degrees of immunosuppression. EBVaGC is usually accompanied by more CD8+ cytotoxic T lymphocytes (CTLs) and mature dendritic cells than EBVnGC. In addition, immune escape caused by regulatory T cells and immunosuppressive (M2) macrophages has been considered to contribute to the prognosis of gastric cancer. Infiltrating M2 macrophages can inhibit T cell functions such as PD-L1, leading to immune tolerance and poor prognosis.

[0006] Therefore, in-depth biomarkers of immune responses in EBVaGC are very important for further understanding the effective mechanisms of immunotherapy. Summary of the invention

[0007] In view of this, one of the objects of the present invention is to provide a prognostic marker for EBV-related malignant tumors.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] Prognostic markers for EBV-related malignant tumors, the prognostic markers comprising any one or more of the following genes: DUOXA2, CA9, COL2A1, SOWAHA, RGN, APOA1, CKB, TPD52L1, WFDC2, TM4SF4, HMGCS2, SLC15A1, ENTPD8, PCOLCE2, FNDC5, PPP1R9A, MMP7, PLA2G7, NPSR1, APOC1, CXCL9, CHI3L1, IDO1, PLA2G2A, CXCL10.

[0010] Furthermore, the EBV-related malignant tumor is EBVaGC.

[0011] Furthermore, the prognostic markers include any one or more of ENTPD8 and PCOLCE2.

[0012] Furthermore, the Gene ID of ENTPD8 is 377841; the Gene ID of PCOLCE2 is 266577.

[0013] Furthermore, the ENTPD8 was negatively correlated with macrophage infiltration; and the PCOLCE2 was positively correlated with macrophage infiltration.

[0014] A second object of the present invention is to provide an EBVaGC prognosis prediction model constructed based on the aforementioned prognostic markers.

[0015] To achieve the above object, the present invention adopts the following technical solutions:

[0016] The EBVaGC prognostic prediction model constructed based on the aforementioned prognostic markers, wherein the risk score of the EBVaGC prognostic prediction model = (0.0306)*APOA1+(0.0565)*ENTPD8+(0.116)*PCOLCE2.

[0017] The third object of the present invention is to provide a reagent for detecting the expression level of the aforementioned prognostic markers for use in preparing a product for evaluating the prognosis of EBVaGC and / or predicting the efficacy of immunotherapy.

[0018] A fourth object of the present invention is to provide an application of the aforementioned EBVaGC prognosis prediction model in the preparation of a product for evaluating the prognosis of EBVaGC.

[0019] A fifth object of the present invention is to provide an expression promoter of ENTPD8 and / or PCOLCE2 for use in the preparation of a drug for inhibiting the development of EVBaGC and / or improving the efficacy of EVBaGC immunotherapy.

[0020] To achieve the above object, the present invention adopts the following technical solutions:

[0021] Use of an expression promoter of ENTPD8 and / or PCOLCE2 in the preparation of a drug for inhibiting the development of EVBaGC and / or improving the efficacy of EVBaGC immunotherapy.

[0022] Furthermore, the expression promoter of ENTPD8 and / or PCOLCE2 is bleomycin.

[0023] Furthermore, the expression promoter is used in the preparation of a drug for inhibiting EVBaGC cell proliferation and clone formation.

[0024] Preferably, the bleomycin increases the expression levels of ENTPD8 mRNA and protein; the bleomycin increases the PCOLCE2 protein level.

[0025] Preferably, the immunotherapy is anti-PD-1 / PD-L1 immunotherapy.

[0026] The beneficial effects of the present invention are:

[0027] Based on the transcriptome sequencing data of normal tissues, gastric cancer tissues, EBVaGC and EBVnGC clinical samples in GEO (gene expression omnibus) and TCGA (cancer genome atlas), the present invention screened differentially expressed genes (DEGs) and performed corresponding functional enrichment analysis to determine the gene set associated with gastric cancer occurrence and EBV infection. Then, a prognostic model based on 25 repeated DEGs was constructed and immune correlation was evaluated; the key genes (ENTPD8 and PCOLCE2) identified as involved in EBVaGC formation and immune response regulation will be considered as biomarkers for evaluating immunotherapy response. New markers may also be used to treat EBV-related malignancies by developing new immunotherapies. Subsequently, the present invention analyzed the interaction between BLM and biomarkers using three-dimensional (3D) protein modeling and molecular docking analysis. The results showed that 572 down-regulated genes and 162 up-regulated genes were detected in normal tissues relative to GC tissues; 196 down-regulated genes and 240 up-regulated genes were detected in EBVnGC relative to EBVaGC, and log2FC≥2, p value≤0.05. Among them, the reduction of ENTPD8 and PCOLCE2 in EBVaGC was significantly associated with prognosis and negatively mediated immune response dysregulation. In addition, when combined with bleomycin, the expression trends of ENTPD8 (positive) and PCOLCE2 (negative) were also opposite, and the binding energy was the most stable, at -4.589kcal / mol and -4.025kcal / mol, respectively. In summary, the present study found that ENTPD8 and PCOLCE2 are specific biomarkers for EBVaGC, and the improvement of immunotherapy by bleomycin may depend on the fluctuation of EBVaGC intratumoral immune response and cell proliferation mediated by ENTPD8 and PCOLCE2 expression. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Figure 2 shows the GSEA items that are significantly enriched in normal tissues and GC tissues. Lines of different colors in the figure represent different pathways, and the gray curve represents the ranking of pathway-related genes based on expression.

[0029] Figure 2 Figure 2 shows the GSEA items that are significantly enriched in EBVnGC and EBVaGC. Lines of different colors in the figure represent different pathways, and the gray curve represents the ranking of pathway-related genes based on expression.

[0030] Figure 3 -A is a bar graph showing the relative proportion of infiltrating immune cells in all samples, where each color represents a type of immune cell; Figure 3-B is an immune score heat map showing 23 activity scores of anti-cancer immunity in the 7-step cancer-immunity cycle, where each row represents an immune state and the gene numbers annotated in the references are shown in the row names.

[0031] Figure 4 Heat map of DEG expression screened for normal tissues and GC tissues, with the clinical information of the samples arranged on the top and the labels listed on the right.

[0032] Figure 5 Heat map of DEG expression screened for EBVnGC and EBVaGC, with the clinical information of the samples arranged on the top and the labels listed on the right.

[0033] Figure 6 Bubble charts are used for the top 10 BP terms in normal tissues and GC tissues, and the size and color of the circles represent the gene count and p-value of each term.

[0034] Figure 7 Bubble plots are used for the top 10 MF terms in normal tissues and GC tissues. The size and color of the circles represent the gene count and p-value of each term.

[0035] Figure 8 Bubble charts are used for the top 10 CC terms in normal tissues and GC tissues, and the size and color of the circles represent the gene count and p-value of each term.

[0036] Fig. 9 Bubble charts are used for the top 10 KEGG terms in normal tissues and GC tissues, and the size and color of the circles represent the gene count and p-value of each term.

[0037] Fig.10 Bubble charts are used for the top 10 BP terms in normal tissues and GC tissues, and the size and color of the circles represent the gene count and p-value of each term.

[0038] Fig.11 Bubble plots are used for the top 10 MF terms in normal tissues and GC tissues. The size and color of the circles represent the gene count and p-value of each term.

[0039] Fig.12 Bubble charts are used for the top 10 CC terms in normal tissues and GC tissues, and the size and color of the circles represent the gene count and p-value of each term.

[0040] Fig.13 Bubble charts are used for the top 10 KEGG terms in normal tissues and GC tissues, and the size and color of the circles represent the gene count and p-value of each term.

[0041] Fig.14 The Venn diagrams show the double down-regulated and up-regulated genes in normal tissues and GC tissues, and EBVnGC and EBVaGC.

[0042] Fig.15 This is a histogram of the expression levels of 25 repeated genes drawn based on sequencing data. In the figure, black represents normal tissue and gray represents tumor tissue.

[0043] Fig.16 This is a histogram of the expression levels of 25 repeated genes drawn based on sequencing data. In the figure, gray represents EBV(-) gastric cancer tissue, and black represents EBV(+) gastric cancer tissue.

[0044] Fig.17 Figure 2 is a heat map of the correlation between 25 repeated genes in normal tissues and GC tissues. Red circles and blue circles indicate positive and negative correlations in gene expression, respectively. The size and color of the circles indicate the degree and significance of the correlation, respectively.

[0045] Fig.18 This is a heat map of the correlation between 25 repeated genes in EBVnGC and EBVaGC. The red and blue circles indicate positive and negative correlations in gene expression, respectively. The size and color of the circles indicate the degree and significance of the correlation, respectively.

[0046] Fig.19 -A is a checkerboard diagram of KEGG projects significantly enriched in normal tissues and GC tissues, and the panel shows the related genes; Fig.19 -B is a checkerboard diagram of KEGG items significantly enriched in EBVnGC and EBVaGC, and the panels show the related genes.

[0047] Fig. 20 This is a graph showing the results of 25 GC repeat gene analysis based on TCGA, where: Fig. 20 -A is a bubble chart showing the fold change of expression of each gene in normal tissues and GC tissues; Fig. 20 -B is a bubble chart of the expression fold changes of each gene in different subtypes; Fig. 20 -C is the trend graph of gene expression from the first stage to the fourth stage; Fig. 20 -D is a plot of the percentage of cancers in which a gene has an effect on the pathway in the selected cancer type (FDR≤0.05), with the number in each cell representing the percentage.

[0048] Fig.21 Figure 2 is the experimental result diagram of prognostic pattern related to 25 repeated DEGs, among which, Fig.21 -A is a graph that uses the lambda parameter to represent the coefficient of the selected feature. The horizontal axis represents the value of lambda and the vertical axis represents the coefficient of the independent variable. Fig.21 -B is the relationship between partial likelihood deviation and log(λ) drawn using the LASSOCox regression model; Fig.21-C is the risk score diagram of the selected data set. The top scatter plot indicates the risk score from low to high; different colors represent different groups; the scatter plot distribution indicates the Riskscore corresponding to the survival time and survival status of different samples; the bottom heat map shows the expression of PCOLCE2, ENTPD8 and APOA1; Fig.21 -D is a graph of Kaplan-Meier survival analysis among groups using log-rank test, in which HR (high exp) represents the hazard ratio of low expression samples relative to high expression samples; Fig.21 -E is the ROC curve and AUC plot of the gene.

[0049] Fig. 22 The overall survival curve of ENTPD8 and PCOLCE2 in gastric cancer patients, among which, Fig. 22 -A is the overall survival curve of gastric cancer patients with ENTPD8, Fig. 22 -B Overall survival curve of PCOLCE2 in gastric cancer patients; the red line represents samples with higher gene levels, and the blue line represents samples with lower gene levels.

[0050] Fig.23 Box plots showing the mRNA expression of ENTPD8 and PCOLCE2 in normal tissues and GC, where Fig.23 -A is a box plot showing the mRNA expression of ENTPD8 in normal tissues and GC; Fig.23 -B is a box plot showing the mRNA expression of PCOLCE2 in normal tissues and GC.

[0051] Fig.24 Immunohistochemical images of normal tissues and GC tissues (n=2) stained with ENTPD8 and PCOLCE2 antibodies (target proteins are in brown).

[0052] Fig.25 : is a graph showing the expression of ENTPD8 and PCOLCE2 in different molecular subtypes (CIN, EBV, GS and MSI); wherein, Fig.25 -A is a diagram showing the expression of ENTPD8 in different molecular subtypes; Fig.25 -B is a graph showing the expression of PCOLCE2 in different molecular subtypes.

[0053] Fig.26 is a graph showing the expression of ENTPD8 and PCOLCE2 in stages I to IV; wherein, Fig.26 -A is a graph showing the expression of ENTPD8 in stages I to IV; Fig.26 -B is a graph showing the expression of PCOLCE2 in stages I to IV.

[0054] Fig. 27A graph scoring the pathway activation of ENTPD8 and PCOLCE2 in the cell cycle, EMT, hormone ER, hormone AR, and apoptosis pathways; where, Fig. 27 -A is a graph showing pathway activation scores of ENTPD8 and PCOLCE2 in the cell cycle pathway; Fig. 27 -B is a diagram showing the pathway activation scores of ENTPD8 and PCOLCE2 in the EMT pathway; Fig. 27 -C is a graph showing pathway activation scores of ENTPD8 and PCOLCE2 in the hormone ER pathway; Fig. 27 -D is a graph showing the pathway activation scores of ENTPD8 and PCOLCE2 in the hormone AR pathway; Fig. 27 -E is a graph showing the pathway activation scores of ENTPD8 and PCOLCE2 in the apoptosis pathway.

[0055] Fig.28 Figure 3 is a graph of immune scores for ENTPD8 and PCOLCE2 expression; wherein, Fig.28 -A is the expression of ENTPD8 in GC samples of the non-responder group (Non-Responder (NR)) and the responder group (Responder (R)); Fig.28 -B is the expression of PCOLCE2 in GC samples of non-responder group (Non-Responder (NR)) and responder group (Responder (R)); Fig.28 -C is a heat map of the correlation between ENTPD8 and PCOLCE2 levels and various immune indicators, in which red boxes indicate positive correlations and blue boxes indicate negative correlations; Fig.28 -D is a graph showing the correlation between macrophage infiltration and prognosis of gastric cancer, in which the red curve represents a higher level of infiltration and the blue curve represents a lower level of infiltration; Fig.28 -E is a graph showing the correlation between macrophage infiltration and ENTPD8 expression, each point represents a GC sample, and the black line indicates the correlation coefficient; Fig.28 -F is a graph showing the correlation between macrophage infiltration and PCOLCE2 expression, each point represents a GC sample, and the black line represents the correlation coefficient.

[0056] Fig.29 The molecular docking analysis results of BLM in the three-dimensional structure of ENTPD8 and PCOLCE2 are shown in Figure 1. Fig.29 -A is a bubble chart of the correlation between BLM and target genes. In the figure, blue bubbles represent negative correlation, red bubbles represent positive correlation, bubble size is positively correlated with FDR significance, and black borders indicate FDR≤0.05; Fig.29 - B is a diagram of the two-dimensional chemical formula of BLM; Fig.29 -C is the three-dimensional structure modeling diagram of ENTPD8; Fig.29-D is the three-dimensional structure modeling diagram of PCOLCE2; Fig.29 -E is a diagram of the first three stable binding modes of BLM and ENTPD8, where the dotted lines represent hydrogen bonds; Fig.29 -F is a diagram of the first three stable binding modes of BLM and PCOLCE2, where the dotted lines represent hydrogen bonds.

[0057] Fig.30 The graph shows the experimental results of the relative expression of ENTPD8 mRNA or protein in SNU719 cell line treated with 50 nM BLM (24 h).

[0058] Fig.31 for Fig.30 Statistical graph of experimental results.

[0059] Fig.32 The graph shows the experimental results of the relative expression of PCOLCE2 mRNA or protein in SNU719 cell line treated with 50 nM BLM (24 h).

[0060] Fig.33 for Fig.32 Statistical graph of experimental results.

[0061] Fig.34 This is a graph showing the results of the CCK-8 experiment.

[0062] Fig.35 This is a graph showing the results of the colony formation experiment. DETAILED DESCRIPTION

[0063] The technical solution of the present invention will be further described clearly and completely in conjunction with specific embodiments below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Therefore, based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work belong to the protection scope of the present invention.

[0064] The Gene ID of ENTPD8 in the NCBI database is 377841. In the TME, immune cells release extracellular adenosine triphosphate (ATP) and drive immune responses through P2X and P2Y receptors. The most important ectonucleosidases for nucleotide dephosphorylation under physiological conditions are members of the ectonucleoside triphosphate diphosphohydrolase (E-NTPDase) family. They are ubiquitous enzymes that hydrolyze nucleoside triphosphates and diphosphates, and they have an additional plasma-plasma orientation. This ectonucleotidase family consists of 8 members (NTPDase-1 to -8). NTPDase-8, located on the plasma membrane, is the most recently discovered member and is also characterized by 2 trans-plasma membrane domains and an active site facing the extracellular environment. ENTPD8 can initiate many cellular responses, such as scavenging extracellular ATP, which is required to prevent innate intestinal inflammation by inhibiting immune cell glycolytic metabolism.

[0065] PCOLCE2 has a Gene ID of 266577 in the NCBI database. Fibrotic deposition of collagen is a major factor affecting immune cell infiltration within tumors. Collagen is a soluble procollagen molecule with an NH2- (N) and a COOH (C)-terminal propeptide attached. Removal of the propeptide is thought to be required for the efficient integration of soluble collagen into insoluble collagen fibrils, specifically enzymatic cleavage of the C-terminal propeptide by bone morphogenetic protein (BMP) 1. The proteolytic activity of BMP1 is enhanced by two extracellular glycoproteins, procollagen C protease-enhancing protein 1 (PCPE1) and PCPE2 (PCOLCE2). PCOLCE2 is expressed from a sequence tag in the chromosomal region 3q21-q24 and encodes a protein with a domain structure similar to PCPE1, with an amino acid sequence homology of 43%. PCPE proteins appear to control a variety of biological events, including matrix assembly, collagen cross-linking, and cell adhesion and migration.

[0066] Bleomycin (BLM) is a metalloglycopeptide antibiotic produced by actinomycetes, which shows antitumor activity in a variety of malignancies, including lymphoma, germ cell tumors, head and neck cancer, cervical, vulvar, and gastric cancer. The complex of BLM and iron is embedded in DNA, causing single-strand and double-strand breaks in DNA, leading to apoptosis of tumor cells. The cytotoxic activity of BLM is through its ability to produce free radical oxygen species (ROS) to oxidize the deoxyribose of thymidylate and other nucleotides, resulting in single-strand and double-strand breaks in DNA, chromosomal aberrations, gaps, fragments, and translocations. Then, BLM leads to prolonged cell cycle arrest, apoptosis, and mitotic cell death through tumor signaling pathways such as PI3K-Akt, p53, and PPAR. Existing studies have shown that SMAD3 activity is altered with BLM binding, which subsequently triggers the SMAD cascade reaction and inhibits GC progression. In addition, BLM can induce ROS-mediated ER stress and autophagy, leading to the surface exposure of chaperone proteins (calreticulin) and the release of ATP, thereby inducing immunogenic cell death, which is dependent on the infiltration of CD8+ T lymphocytes recruited by dendritic cells (DCs), the secretion of interferon-γ, and the expansion of Foxp3+ regulatory T cells (Tregs) induced by TGF-β.

[0067] In the embodiment of the present invention, the RNA sequence analysis of GC tissue is as follows:

[0068] Transcriptome sequencing dataset GSE51575 was obtained from the GEO database. In the present invention, fresh GC tissues from 203 patients who underwent laparotomy before chemotherapy were cryopreserved. Informed written consent was obtained from the patients, and the protocol was approved by the SMC Institutional Review Board (IRB file number 2010-08-159). Conventional in situ hybridization of EBV-encoded small RNA (EBER) was used to classify EBVaGC or EBVnGC. After microscopic analysis of tumor necrosis, size, margin, and H&E staining, a total of 14 EBVnGCs and 12 EBVaGCs with qualified RNA quality were selected for this study, as shown in Table 1. Total RNA from normal (N) and tumor (T) tissues of each sample was extracted and separated using Trizol (Invitrogen). Whole transcriptome sequencing (Hi RNAseq) was performed using an Illumina HiSeq 2000 instrument. The Agilent G4851A platform (Agilent Technologies; CA, USA) was used for subsequent RNA array analysis. Gene expression data were then analyzed using R language.

[0069] Table 1. Clinical information of samples

[0070] Organization Type The number of samples of each type Organization Type Normal gastric tissue (n=26); gastric tumor (n=26) EBV type EBV(-) tumors (n=14); EBV(+) tumors (n=12) Histology CA tumors (n=17); CLR tumors (n=6); LELC tumors (n=3) Lauren Type Intestinal (n=13); Diffuse (n=8); Mixed (n=3); Uncertain (n=2) age 50~60(n=12); 60~70(n=20); 70~80(n=12); Others(n=8) gender Male (n=42); Female (n=10) Tumor stage Phase I AC (n=1); Phase II AC (n=10); Phase III AC (n=10) Her2+ Type Her2+:-(n=49); Her2+:1+(n=2); Her2+:1+++(n=1)

[0071] The quantile method was used for background removal and quantile normalization. The RNAseq sequences were aligned to the human reference genome (hg19) using the STAR alignment tool. The empirical Bayes method was used to detect DEGs in normal tissues versus tumor tissues, and EBVnGC versus EBVaGC. The Benjamini-Hochberg procedure was used to adjust the p-value for multiple calibration. The differentially expressed genes were screened for p-value ≤ 0.05 and logFC ≥ 1.0. Functional enrichment analysis and visualization of differentially expressed genes were performed using DAVID and ClueGO. Enriched BP: biological process; CC: cellular component; MF: molecular function and pathway KEGG were determined. One set of significant DEGs was further applied to the BiNGO tool of Cytoscape to visualize the gene ontology hierarchy and functional interaction graphs between significant DEGs. Table 2 lists some of the other databases used.

[0072] Table 2. Database list

[0073] Database ID URL GEOdatabase https: / / www.ncbi.nlm.nih.gov / gds / ?term= TCGA https: / / www.cancer.gov / TISIDB http: / / cis.hku.hk / TISIDB / index.php The Human Protein Atlas https: / / www.proteinatlas.org / Expasy http: / / web.expasy.org / protparam / PSIPRED http: / / bioinf.cs.ucl.ac.uk / psipred / DNMIVD http: / / 119.3.41.228 / dnmivd / index / GSCALite http: / / bioinfo.life.hust.edu.cn / web / GSCALite / LinkedOmics http: / / www.linkedomics.org / STRING https: / / string-db.org / DrugBank https: / / go.drugbank.com / SWISS-MODEL https: / / swissmodel.expasy.org / interactive#alignment GEPIA http: / / gepia.cancer-pku.cn / index.html AutodockVina1.2.2 http: / / autodock.scripps.edu / TIP http: / / biocc.hrbmu.edu.cn / TIP / index.jsp

[0074] In the examples of the present invention, the cell lines and culture conditions are as follows:

[0075] The EBVaGC cell line (SNU719) was used in this study. SNU719 was obtained from the American Type Culture Collection (ATCC). Cell lines were routinely tested for mycoplasma contamination. All cell lines were cultured in RMI-1640 medium (Gibco, Carlsbad, CA, USA) containing 10% heat-inactivated fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA) and 2% penicillin-streptomycin at 37°C and 5% CO2.

[0076] In the embodiment of the present invention, the RNA separation and real-time quantitative PCR method are as follows:

[0077] Total RNA was extracted from the cell lines using TRIzol reagent (Ambion, USA) and then reverse transcribed using the First Strand cDNA Synthesis Kit (TaKaRa, Japan). Then, the cDNA was used as a PCR template and the mRNA level of the target gene was detected using the FastStart DNA Master SYBR Green Kit (Roche Diagnostics, Mannheim, Germany). PCR conditions were set according to the manufacturer's protocol. All reactions were performed in triplicate, and the Ct values ​​of GAPDH or U6 were used for normalization purposes. Each experiment included 3 technical replicates. Two -△△Ct The primers used in the present invention are shown in Table 3.

[0078] Table 3. PCR primer list

[0079]

[0080] In the embodiment of the present invention, the Western blotting method is as follows:

[0081] All cells were lysed with radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitor phenylmethylsulfonyl fluoride (PMSF) and phosphatase inhibitor cocktail according to the manufacturer's instructions. Protein concentration was quantified using bis-cholic acid (BCA) protein assay kit (CWBIO, Beijing, China). Protein extracts were separated by 10% SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (0.22 μm, Millipore, Germany). The membranes were blocked with 5% skim milk for 2 h at room temperature and then incubated with primary antibodies (Anti-ENTPD8 antibody, Beyotime Biotechnology; anti-PCOLCE2 antibody (ab232721) and anti-GAPDH antibody (ab8245)) at 4 °C overnight and with secondary enzyme-labeled antibodies (ZSBIO, Beijing, China) for 2 h at room temperature. Immunoreactive bands were detected using Immobilon ECL substrate kit (Millipore, Germany). Images were acquired using a chemiluminescence system (Bio-Rad, USA). Densitometry values ​​of target bands were normalized with GAPDH to quantify protein content. Three biological and technical replicates were included in each experiment. Grayscale analysis of Western blots was performed using ImageJ.

[0082] In the embodiment of the present invention, the cell proliferation (CCK-8) determination method is as follows:

[0083] Cell proliferation CCK8 assay was performed using the Cell Counting Kit-8 (CCK8; Boster Biotech Co., Ltd., Wuhan). 5000 SNU719 cells were collected from the culture dish and transferred to a new sterile 96-well plate. An appropriate amount of culture medium was added to each well to ensure that all cells were in contact with the culture medium. After 24 h of incubation, 50 nM bleomycin was added to the experimental wells and an equal amount of buffer was added to the control wells. The cells were grown for another 1, 2, 3 or 4 days. After the selected time point, 100 μl of the diluted CCK-8 (10%) reagent was transferred to each well of the 96-well plate and incubated at 37 °C for another 4 h. During the incubation period, the CCK-8 reagent will be reduced to fumaric acid, a purple fumarate. The absorbance was measured at 450 nm using a microplate reader. The average absorbance of each sample was calculated. The average absorbance values ​​were compared to determine the number of viable cells in the sample. Each experiment included 3 biological replicates and 6 technical replicates.

[0084] In the embodiment of the present invention, the clone formation test method is as follows:

[0085] Cells were plated at a density of 1000 cells per well in 6-well plates and maintained in RPMI-1640 medium supplemented with 10% FBS. After 15 days, cells were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet for 20 min at room temperature. Crystal violet is an important stain that binds to the extracellular matrix and can be used to quantify cell viability and proliferation. Once staining is complete, the plates are washed with PBS to remove any excess stain and then photographed. The number of colonies with more than 50 cells is counted.

[0086] In the embodiment of the present invention, the survival curve analysis method is as follows:

[0087] Overall survival analysis was performed using the R package survival, and patients were dichotomized according to median expression. Survival curves were constructed using the Kaplan-Meier survival estimator. The overall survival of the two groups of patients was compared using the Log-rank test (corresponding to the two-sided z test), and the hazard ratio (HR) (95% confidence interval) for the comparison of the two groups was provided. Multiple tests were performed according to the Benjamini-Hochberg method, and the p value was adjusted according to the false discovery rate (FDR). The proportional hazards assumption was tested.

[0088] In the embodiment of the present invention, the GC gene expression and function analysis method based on the TCGA database is as follows:

[0089] Based on normalized and batch-corrected RSEM mRNA expression, differential expression analysis of STAD is provided with more than ten paired tumor and normal samples. Fold change was calculated using mean(Tumor) / mean(Normal), and p-values ​​were estimated using t-tests with further FDR adjustment.

[0090] For expression and survival analysis, clinical data were used, while samples with competing mortality risk were filtered out. The mRNA expression and related clinical survival data were merged and divided into high expression group and low expression group based on the mRNA median. The survival time and survival status of the two groups were then fitted using the R package survival. Log-rank test was performed for each gene in GC. In the analysis of expression and subtype or stage, mRNA expression was merged with clinical stage data by sample barcode. There must be at least 5 samples in the subtype and stage groups. Pathological stage and clinical stage divided the samples into stage I, stage II, stage III and stage IV. The Mann-Kendall trend test was used for stage trend analysis.

[0091] The correlation between expression and pathway activation was estimated by the difference in gene expression between the pathway activation group and the inhibition group, defined by the median pathway score. The GC pathway activity score was calculated using the RPPA (Reverse phase protein array) data in the TCPA database. Well-known cancer-related pathways include TSC / mTOR, RTK, RAS / MAPK, PI3K / AKT, hormone ER, hormone AR, EMT, DNA damage response, cell cycle, and apoptosis pathways. The samples were divided into high and low groups according to the median gene expression, and the differences in pathway activity scores were tested using a t-test. The p-value was corrected by FDR and was significant when FDR≤0.05. Once score (high level)>score (low level), gene A may activate this pathway, otherwise it has an inhibitory effect.

[0092] In the embodiment of the present invention, the molecular docking of bleomycin and protein is as follows:

[0093] In order to analyze the binding affinity and interaction mode of candidate drugs with targets, the protein-ligand docking software AutodockVina1.2.2 was used. The molecular structure of BLM (DrugBank retrieval number: DB00290) was retrieved from the DrugBank database. Due to limited research on ENTPD8 and PCOLCE2 proteins, there are no fixed PDB files, and their three-dimensional structures are unclear. The three-dimensional coordinates of ENTPD8 and PCOLCE2 were established based on the amino acid sequences using the SWISS-MODEL software, and the PDB files were downloaded for docking. All protein and molecular files were converted to PDBQT format, all water molecules were excluded, and polar hydrogen atoms were added. The grid box is located in the center, covering the area of ​​each protein and accommodating free molecular motion. The grid box is set to The grid point distance is 0.05 nm. Molecular docking study was performed by Autodock Vina1.2.2, and the top 9 results of binding energy are listed.

[0094] In the embodiment of the present invention, the statistical data are as follows:

[0095] Statistical analysis was performed using SPSS software for Windows V.13.0 (SPSS Inc., Chicago, IL, USA) and R version 3.4.0. The chi-square test was used to evaluate the correlation between ENTPD8 and PCOLCE2 levels by immunohistochemical staining and clinicopathological characteristics. Univariate and multivariate COX regression analysis was used to evaluate the correlation between ENTPD8 and PCOLCE2 levels and prognosis. Survival curves were drawn using the Kaplan-Meier method and log-rank test. KEGG pathway and GO enrichment analysis were calculated using the R package “clusterProfile” based on the gene expression matrix. The statistical assumptions in this paper were all two-tailed, and p values ​​≤ 0.05 were considered significant.

[0096] Example 1. Analysis of immune system dysregulation in EBVaGC

[0097] A transcriptome sequencing dataset (GSE51575) was obtained from the GEO database to analyze the mechanisms and characteristics of EBVaGC. A total of 52 clinical samples were collected in this study, including 26 pairs of GC tissues and corresponding tumor-adjacent tissues; all tumors were divided into EBVnGC (n=14) and EBVaGC (n=12). The quality control results showed that the sequencing accuracy was similar, and two absorption peaks were found at intensity 5 and intensity 10 in the expression density maps of all samples. The PCA dimensionality reduction clustering results based on the corrected data showed that normal tissues and cancer tissues could be divided into two groups. Although there were slight differences between EBCnGC and EBVaGC, their spatial distribution was relatively close. In order to screen the key molecules and molecular mechanisms of EBVaGC carcinogenesis, the differences in transcriptional profiles between normal tissues and GC tissues, and between EBVnGC and EBVaGC were compared, and DEGs were identified with logFC≥1 and p value≤0.05, and the gene symbols of the top 20 significant DEGs were marked.

[0098] Abnormal gene transcription often leads to dysregulation (activation or inhibition) of intracellular signaling pathways, accompanied by significant changes in biological functions. To this end, gene set enrichment analysis (GSEA) was performed on the whole gene transcripts of normal tissues and GC tissues or EBVnGC and EBVaGC. Figure 1 , Figure 2 , Figure 3 -A. Figure 3-B shows that 15 significantly enriched pathways (p-value < 0.05) in normal tissues compared with GC tissues are almost all related to energy metabolism, such as fatty acid metabolism (p = 0.0061), glycine serine and threonine metabolism (p = 0.0106) and pentose phosphate pathway (p = 0.0378). Most of the genes involved in these pathways are downregulated in gastric cancer, which may indicate that energy metabolism is inhibited, which is consistent with the metabolic reprogramming theory of cancer. In addition, like colorectal cancer, several cancer pathways are enriched. And the energy metabolism pathways of EBVaGC are significantly changed with gene downregulation. In the prior art, Wang et al. characterized EBVaGC-specific cellular pathways and detected changes in the biosynthesis processes of macromolecules such as carbohydrates, lipids and proteins. In EBVaGC, deregulation of cholesterol transport and lipoprotein clearance pathways is also obvious. The present invention further found that after EBV infection, in addition to the enrichment of pathogenic microbial pathways, 10 significantly enriched immune pathways (p < 0.05) were detected for EBVnGC and EBVaGC, such as primary immunodeficiency (p value = 0.006), intestinal immune network producing IGA (p value = 0.0178) and antigen processing and presentation (p value = 0.0401), see for details. Figure 2 The enrichment scores suggest that immune regulatory mechanisms are active in EBVaGC.

[0099] Different intratumoral immune subtypes are often associated with the progression and treatment of gastric cancer. Therefore, the immune scores of all 52 samples were calculated within 7 steps: cancer antigen release, cancer antigen presentation, priming and activation, immune cell trafficking to tumors, immune cell infiltration into tumors, T cell recognition of cancer cells, and cancer cell killing. Figure 3 -B. The immune scores of EBVaGC at each step were almost higher compared to EBVnGC. In addition, the proportion of immune cells was shown in the histogram according to the specific biomarkers, see Figure 3 -A. Among them, CD8+ Cells decreased, while CD8+ memory cells, CD8+ effector cells, and Treg cells increased.

[0100] These results suggest that EBVaGC subtypes may have more aggressive immune responses, leading to better prognosis and sensitivity to immunotherapy.

[0101] Example 2. Enrichment analysis of DEGs in normal tissues and GC tissues and comparison of EBVnGC and EBVaGC

[0102] To identify key biomarkers involved in immune regulation, DEGs in normal tissues and GC tissues, EBVnGC and EBVaGC were screened with logFC ≥ 2.0 and p value ≤ 0.05. Figure 4-5 As shown in Figure 2, a total of 734 DEGs were detected in the comparison between normal tissues and GC tissues (572 down-regulated and 162 up-regulated), and a total of 436 DEGs were detected in EBVnGC and EBVaGC (196 down-regulated and 240 up-regulated), and the DEG expression trend was not significantly correlated with clinical indicators such as age and gender. Then, the two groups of DEGs were enriched and analyzed, as shown in Figure 2. Figure 6 to Figure 13 As shown, the top 10 terms of BP, MF, CC, and KEGG pathways were listed. DEGs in GC may affect gastric digestion (BP), gastric acid secretion (KEGG), and collagen-containing extracellular matrix function (CC). However, DEGs between EBVnGC and EBVnGC were enriched in multiple immune-related functions, such as regulation of T cell activation (BP), antigen processing and presentation of exogenous peptide antigens (BP), MHC class II receptor activity (MF), and IgA production by the intestinal immune network (KEGG). The DEG enrichment results were similar to the above-mentioned GSEA of the whole gene, further showing the immune importance of DEGs in EBVaGC subtypes.

[0103] Example 3. Repeated DEGs in normal tissues, GC tissues, EBVnGC and EBVaGC

[0104] It is well known that EBVaGC is a specific subtype with abnormal immune regulation. To classify the key markers related to tumorigenesis and immune response after exposure to EBV, a total of 25 repeated DEGs (16 double down-regulated and 9 double up-regulated) and transcriptome sequencing data of all genes were identified (down-regulated: DUOXA2, CA9, COL2A1, SOWAHA, RGN, APOA1, CKB, TPD52L1, WFDC2, TM4SF4, HMGCS2, SLC15A1, ENTPD8, PCOLCE2, FNDC5, and PPP1R9A; up-regulated: MMP7, PLA2G7, NPSR1, APOC1, CXCL9, CHI3L1, IDO1, PLA2G2A, and CXCL10), as detailed in Figure 14 to Figure 16 In addition, DEGs had higher correlations in normal tissues and GC tissues compared with EBVnGC and EBVaGC, as detailed in Figure 17-18 , which may indicate that the repeated DEGs may share certain regulatory mechanisms and functions in GC formation, while the correlation is reduced due to the possibility of transcriptional abnormalities caused by EBV infection as an exogenous variable. To further understand the functions of repeated DEGs in EBVaGC, enrichment analysis of down-regulated and up-regulated genes was performed again, and the results are shown in Fig.19As shown in the figure. In the KEGG results, metabolism-related pathways (metabolic pathways, p value = 0.0016) and immune-related pathways (cytokine-cytokine receptor interaction, p value = 0.0167) were the most significant. The enrichment function of repeated DEGs was also similar to the above GSEA results, indicating that these 25 genes as a whole play a key regulatory role in the formation and immune response of EBVaGC.

[0105] Example 4. Role of repeated DEGs in cancer-related pathways

[0106] Although the 25 duplicate genes screened have been confirmed to mediate metabolic and immune-related pathways, which have been shown to play a key role in gastric cancer, it is necessary to establish the association between DEGs and cancer-related pathways. Based on the mRNA expression profiles and clinical information of GC samples (n = 380), the fold changes of DEGs in normal tissues compared with GC tissues and different subtypes were calculated, and the results are shown in Fig. 20 -A and 20-B, but some genes such as DUOXA did not show obvious expression differences, which may be due to the large number of samples and heterogeneity. Fig. 20 -A shows the expression fold change of each gene in normal tissue and GC tissue. In this figure, logFC and FDR are represented by color and size; this row lists the gene symbols; the bubble color from purple to red represents the fold change between normal and GC; the size of the point is positively correlated with the FDR significance. Fig. 20 -B shows the expression fold changes of each gene in different subtypes. In this figure, the bubble color from white to red represents the importance of FDR; the bubble size is positively correlated with the significance; the black outline border of the bubble indicates FDR≤0.05.

[0107] In addition, a line graph is used to show the expression trend of stages I-IV, see Fig. 20 -C, the color of the trend line in the figure changes from blue to red, indicating a trend from decreasing to increasing. Among them, SOWAHA and ENTPD8 continued to decrease, while CKB and SLC15A1 continued to be induced from stage I to stage IV. Fluctuations in gene levels will directly affect the status of intracellular pathways. In order to characterize the role of duplicated genes in gastric cancer, we calculated the correlation between gene expression and the status of 10 classic pathways involved in cancer development. The results are shown in Figure 2. Fig. 20 -D. For example, the apoptosis pathway was most associated with DEGs. PI3K-AKT (WFDC2, SLC15A1, RGN, and APOA1) and RASMAPK (TM4FS4, RGN, and DUOXA2) pathways were only detected to be activated. Interestingly, both PI3K and MAPK have been shown to mediate the immune system in GC.

[0108] This result suggests that the 25 screened DEGs can not only regulate cancer-related pathways but also mediate the intratumoral immune system.

[0109] Example 5. Prognostic model of 25 repeated DEGs in GC

[0110] In order to estimate the value of this DEG set in the prognosis of GC patients, we constructed a mathematical model based on the gene level using the LASSOCox regression model. In this model, all GC patients were divided into two groups with different risk scores (high and low risk) according to the expression of PCOLCE2, ENTPD8 and APOA1 (lambda.min = 0.0444), as shown in Figure 2. Fig.21 -A~ Fig.21 -C. PCOLCE2 and APOA1 were positively correlated with the risk score, while ENTPD8 was negatively correlated with the risk score. This suggests that the expression levels of PCOLCE2, ENTPD8 and APOA1 are the most important factors affecting the prognosis of this model. The risk score formula is:

[0111] Risk score = (0.0306)*APOA1+(0.0565)*ENTPD8+(0.116)*PCOLCE2

[0112] According to the risk score, the patients were divided into high-risk group and low-risk group (n = 185). The median survival time was 1.7 years and 4.6 years, respectively. Fig.21 -D. Log-rank p value = 5.64e-05, HR (high group) = 2.001, 95% CI (1.428, 2.804). This suggests that the prognostic model can divide patients into two groups with significant differences in survival rates. To evaluate whether the model can objectively predict prognosis, the AUC values ​​of 1 year (0.605, 95% CI (0.545-0.665)), 3 years (0.696, 95% CI (0.618-0.773)), 5 years (0.615, 95% CI (0.48-0.751)), and 9 years (0.661, 95% CI (0.401-0.922)) were calculated. For details, see Fig.21 -E, the higher the AUC value, the stronger the predictive ability. Among them, the 3-year AUC value of the prognostic model was about 0.7, indicating that PCOLCE2, ENTPD8 and APOA1 had good prognostic evaluation effect and clinical reference value.

[0113] Example 6. Expression of ENTPD8 and PCOLCE2

[0114] In the process of constructing the prognostic model, we found that not all provided DEGs could effectively contribute to the risk score of patients, indicating that some of them would not significantly change the overall survival of patients. Therefore, the survival curves of each gene were plotted according to the expression level, and only ENTPD8 (Logrankp value = 0.014) and PCOLCE2 (Logrankp value = 0.00093) could significantly lead to a poor prognosis of gastric cancer in contrast to the expression level, as shown in Figure 2. Fig. 22 Further detection revealed that the levels of ENTPD8 and PCOLCE2 in GC tissues were reduced at both the mRNA level and the protein level. Fig.23 and Fig.24 In addition, according to the TCGA data of GC, ENTPD8 and PCOLCE2 were also downregulated in the EBVaGC subtype, see Fig.25 At the same time, the expression trends of ENTPD8 and PCOLCE2 in clinical stages I, II, III, and IV were not significant, see Fig.26 . The results suggest that both ENTPD8 and PCOLCE2 can be used as specific biomarkers for EBVaGC and affect prognosis. Similarly, by calculating the pathway activation score, it was found that ENTPD8 can activate the cell cycle pathway (p value = 1.5e-02), while inhibiting the EMT pathway (p value = 3.1e-04) and Hormone ER pathway (p value = 3.5e-02). PCOLCE2 can also inhibit the hormone AR pathway (p value = 2.5e-02) and the apoptosis pathway (p value = 2.2e-06), see for details. Fig. 27 .

[0115] Example 7. Immunomodulatory effects of ENTPD8 and PCOLCE2 in gastric cancer

[0116] In this study, the most significant difference between EBVnGC and EBVaGC was the intratumoral immune response, which will determine the efficacy of immunotherapy. According to the anti-PD-1 / PD-L1 GC dataset (PRJEB25780), ENTPD8 (Wilcoxon p value = 0.51) and PCOLCE2 (Wilcoxon p value = 0.19) were both downregulated in the responder group, but there was no statistical significance, see Fig.28 -A and Fig.28 -B. The results suggest that ENTPD8 and PCOLCE2 may serve as targets and biomarkers for EBVaGC immunotherapy. In order to comprehensively study the mechanism of action of ENTPD8 and PCOLCE2 in GC immune regulation, the correlation between their expression levels and various immune events was calculated. The results are shown in Fig.28-C. The results showed that ENTPD8 and PCOLCE2 had an inverse correlation in lymphocyte infiltration and immune factor expression. For example, among all lymphocytes, only macrophage infiltration affected the prognosis (Logrank p value = 0.004), see Fig.28 -D. In addition, by calculating the correlation coefficient, it was shown that ENTPD8 (partial.Cor = -0.299; p-value = 4.60e-09) was negatively correlated with macrophage infiltration; while PCOLCE2 (partial.Cor = 0.34; p-value = 1.88e-11)) was positively correlated with macrophage infiltration, see Fig.28 -E and Fig.28 -F.

[0117] Example 8. BLM binding sites in ENTPD8 and PCOLCE2 protein structures

[0118] The expression levels of ENTPD8 and PCOLCE2 may affect EBVaGC immunity and play a role in anti-PD-1 / PD-L1 immunotherapy. Screening drugs that can regulate the expression or stability of ENTPD8 and PCOLCE2 is expected to serve as adjuvants to improve the effectiveness of immunotherapy and reduce drug resistance. According to previous studies, the correlation between drugs and ENTPD8 and PCOLCE2 levels was calculated. For details, see Fig.29 -A. A total of 21 drugs that simultaneously regulate ENTPD8 and PCOLCE2 levels were identified. Among them, 50 μM BLM was the most effective, with FDR ≤ 0.05. BLM can inhibit DNA metabolism and is used as an anti-tumor drug with a systemic absorption rate of 45%. The molecular formula of BLM is C 55 H 84 N 17 O 21 S3, molecular weight is 1415.552, two-dimensional chemical formula is Fig.29 -B.

[0119] Due to the lack of research on protein structure, ENTPD8 and PCOLCE2 do not have clear three-dimensional conformations. In order to conduct molecular-protein docking analysis, ENTPD8 and PCOLCE2 were modeled based on their amino acid sequences to obtain PDB files describing their spatial structures. Fig.29 -C. Fig.29 To evaluate the affinity of BLM for ENTPD8 and PCOLCE2, we performed molecular docking analysis. AutodockVina v.1.2.2 software was used to obtain binding poses and interactions and generate binding energies for each interaction. The results are shown in Fig.29 -E, Fig.29-F, Table 4, Table 5. The results show that BLM binds to protein targets through visible hydrogen bonds and strong electrostatic interactions. In addition, the five candidate drugs successfully occupied the hydrophobic pocket of each target. For ENTPD8, the BLM target site consists of 4 hydrogen bonds (TRP440, THR125, GLU168, and GLU168), with the lowest binding energy of -4.589 kcal / mol, and the binding is highly stable. Similarly, the most stable binding site of BLM in PCOLCE2 consists of 2 hydrogen bonds (CYS86 and LEU85) with a binding energy of -4.025 kcal / mol.

[0120] Table 4. Molecular docking analysis of the number and position of hydrogen bonds between bleomycin and ENTPD8

[0121] Binding Energy quantity The location of hydrogen bonds -4.589 kcal / mol n＝4 TRP440,THR125,GLU168,GLU168 -4.561 kcal / mol n＝3 ILE438,ASP48,HIS53 -4.524 kcal / mol n＝1 SER52 -4.465 kcal / mol n＝1 ASN354 -4.387 kcal / mol n＝2 SER86,LEU130 -4.326 kcal / mol n＝1 LEU130 -4.304 kcal / mol n＝3 TYR357,TRP440,ASP205 -4.275 kcal / mol n＝3 SER55,HIS53,TYR357 -4.265 kcal / mol n＝2 ASP205,ASP205

[0122] Table 5. Molecular docking analysis of the number and position of hydrogen bonds between bleomycin and PCOLCE2

[0123] Binding Energy quantity The location of hydrogen bonds -4.025 kcal / mol n＝2 CYS86,LEU85 -3.84 kcal / mol n＝2 ASN84,ASN84 -3.633 kcal / mol n＝2 ASP151,SER164 -3.624 kcal / mol n＝4 SER117,VAL116,TYR153,GLY155 -3.612 kcal / mol n＝1 CYS231 -3.61 kcal / mol n＝2 CYS154,PRO146 -3.594 kcal / mol n＝1 TYR212 -3.538 kcal / mol n＝2 LEU85,VAL221 -3.525 kcal / mol n＝0 N / A

[0124] To investigate the role of BLM in EBVaGC, we treated SNU719 cell line with 50 nM BLM and performed CCK-8 and colony formation assays. The results showed that BLM increased the expression levels of ENTPD8 mRNA and protein compared with the control group (p value 0.05). PCOLCE2 protein level increased, but mRNA level did not change significantly. Figure 30 to Figure 33 It is noteworthy that when EBVaGC cells were treated with BLM, cell proliferation and colony formation were significantly reduced, as detailed in Fig.34 and Fig.35 This suggests that bleomycin itself, in addition to being a potential immunotherapy adjuvant, can inhibit the development of EVBaGC.

[0125] In summary, the present invention depicts the transcriptome landscape of tumor paracancerous tissues and tumor tissues containing EBVnGCs and EBVaGCs. An immune-related gene set containing 25 repeated DEGs was identified as a regulatory network for EBVaGC immunotherapy. Existing studies have shown that EBVaGC is an important subtype and is considered to be potentially sensitive to immunotherapy such as ICB (PD-1 / PD-L1) due to a higher frequency of PD-L1 (30%) and a relatively "hot" immune TME compared with EBVnGC. It is speculated that EBVaGC with higher PD-L1 levels are more likely to respond to immunotherapy and are more sensitive to ICB. However, the effect of PD-L1 expression on the prognosis of gastric cancer remains controversial. Once intratumoral T cells are activated, PD-1 is strictly regulated and strongly induced, and can bind to the ligand (PD-L1) to form an immunosuppressive TME. PD-L1+EBVaGC has a deeper invasion, a poorer prognosis, and a higher degree of lymphocyte infiltration. Wei et al., in "The clinicopathological significance and predictive value for immunotherapy of programmed death ligand-1 expression in Epstein-Barr virus-associated gastric cancer", collected and analyzed publicly available clinical data, providing an important perspective on biomarkers for EBVaGC immunotherapy. The ORR in the PD-L1-EBVaGC group was 0%, while the ORR in the PD-L1+EBVaGC group was 63.3%. The PFS in the PD-L1+EBVaGC group was significantly prolonged.

[0126] TME is complex and diverse, and has also been shown to have an important impact on prognosis and immunotherapy efficacy. However, according to some clinical reports, the efficacy of ICB for all EBVaGC may not be universally satisfactory and still faces many challenges. This study analyzed the characteristics of TME in EBVaGC patients based on the expression of ENTPD8 and PCOLCE2. Further deepening the understanding of the role of TME in anti-tumor and immune responses will guide more effective immunotherapy strategies. However, the results of retrospective analysis based on public clinical reports are limited to confirming the role of PD-L1 in EBVaGC immunotherapy. In the present invention, PD-L1 / 2 was negatively correlated with ENTPD8 and positively correlated with PCOLCE2.

[0127] Although CD8+T cell subsets are known to be associated with the mechanism of action of these immunotherapeutics, clinical applications require a comprehensive and diverse set of markers to provide comparable prognostic accuracy. In a spontaneous GC mouse model, macrophage infusion successfully activated and recruited CD8+T cells. In this study, ENTPD8 and PCOLCE2 were also found to be associated with macrophage infiltration, which was negatively correlated with prognosis and could be used as complementary markers. In gastric cancer, macrophages are an important component of the innate immune system. The ratio of pro-inflammatory to anti-inflammatory macrophages in gastric cancer is closely related to survival, because polarization of pro-inflammatory subtypes can lead to apoptosis of gastric cancer cells. In addition, macrophages can secrete CXCL8 and participate in the immunosuppressive microenvironment by inducing PD-L1+ subtypes. At the same time, existing studies have also found that the CD68+STING+ macrophage subtype leads to poor OS and higher density, suggesting that STING or macrophages are negative prognostic factors for GC. Interestingly, CD68+STING+ macrophages were also associated with negative responses to anti-PD-1 / PD-L1 therapy.

[0128] Clinical immunotherapy for EBVnGC still has limitations. Recently, a common solution is to use adjuvant conventional therapy (radiotherapy and chemotherapy) to improve the efficacy, taking into account the ability to induce immunogenic cell death, such as pembrolizumab (MK-3475) with paclitaxel. Similarly, BLM can also induce programmed cell death through the infiltration of CD8+ T lymphocytes. Through molecular docking, the present invention identified the top 9 binding sites of BLM in ENTPD8 and PCOLCE2. Our current study demonstrated the important effect of bleomycin on the immunotherapy of EBVaGC. The results of this study provide important perspectives for future clinical studies on immunotherapy combined with BLM for the treatment of EBVaGC.

Claims

1. A prognostic marker for EBV-related malignant tumors, characterized in that: The prognostic markers include any one or more of the following genes: DUOXA2, CA9, COL2A1, SOWAHA, RGN, APOA1, CKB, TPD52L1, WFDC2, TM4SF4, HMGCS2, SLC15A1, ENTPD8, PCOLCE2, FNDC5, PPP1R9A, MMP7, PLA2G7, NPSR1, APOC1, CXCL9, CHI3L1, IDO1, PLA2G2A, CXCL10.

2. The prognostic marker according to claim 1, characterized in that The EBV-related malignant tumor is EBVaGC.

3. The prognostic marker according to claim 1, characterized in that The prognostic markers include any one or more of ENTPD8 and PCOLCE2.

4. An EBVaGC prognosis prediction model constructed based on the prognostic markers according to any one of claims 1 to 3.

5. The EBVaGC prognosis prediction model according to claim 4, characterized in that: The risk score of the EBVaGC prognostic prediction model = (0.0306)*APOA1+(0.0565)*ENTPD8+(0.116)*PCOLCE2.

6. Use of a reagent for detecting the expression level of the prognostic marker according to any one of claims 1 to 3 in the preparation of a product for evaluating the prognosis of EBVaGC and / or predicting the efficacy of immunotherapy.

7. Use of the EBVaGC prognosis prediction model according to any one of claims 4 to 5 in the preparation of a product for evaluating the prognosis of EBVaGC.

8. Use of an expression promoter of ENTPD8 and / or PCOLCE2 in the preparation of a drug for inhibiting the development of EVBaGC and / or improving the efficacy of EVBaGC immunotherapy.

9. The use according to claim 8, characterized in that: The expression promoter of ENTPD8 and / or PCOLCE2 is bleomycin.

10. The use according to claim 8, characterized in that: The expression promoter is used in preparing a drug for inhibiting EVBaGC cell proliferation and clone formation.

Images