Colorectal cancer tumor border cell subset marker combination and its application

Abstract

The present invention discloses biomarkers of colorectal cancer tumor border cell subpopulations and their applications in diagnosis, prevention, treatment and prognosis. The present invention also discloses a microenvironment cell map of adjacent and cancerous tissues of colorectal cancer patients, in which FAP is specifically enriched in colorectal cancer tissues. + Fibroblasts and SPP1 + The present invention also discloses FAP + Fibroblasts and SPP1 + The present invention further discloses that patients with high expression of these two proteins in bladder cancer data sets also have a lower response rate to PD-L1 treatment. The present invention also discloses FAP + Fibroblasts and SPP1 + Macrophages are also spatially localized in colorectal cancer, forming a "wall"-like tumor boundary around the tumor center, thereby preventing immune cells from entering the tumor center. These cellular characteristics and cell interaction patterns can be used to effectively diagnose, prevent, treat, and assess prognosis and survival for CRC.

Description

Technical Field

[0001] This invention belongs to the field of medical technology and relates to biomarkers of colorectal cancer tumor boundaries and their applications. Through single-cell transcriptional profiling and spatial transcriptome mapping, the authors reveal the roles and interaction patterns of tumor-specific fibroblast and macrophage subsets in tumor boundary formation. Furthermore, by targeting the interaction between these two cell types, they reveal the relationship between markers of colorectal cancer tumor boundaries and disease progression. Background Art

[0002] Recently, immune checkpoint blockade (ICB) strategies have been applied to the treatment of colorectal cancer (CRC). Many studies have explored the use of T cells for effective anti-tumor immunotherapy. However, the PD-1-targeting antibody pembrolizumab is only effective against mismatch repair-deficient tumors with high microsatellite instability (MSI-H), which account for less than 5% of metastatic CRC cases. Therefore, it is necessary to understand the mechanisms of cellular and molecular remodeling in the colorectal cancer tumor microenvironment (TME) and to identify potential intervention targets to enhance the efficacy of immunotherapy. Recent studies have shown that stromal cells and myeloid cells may form unique niches for tumor growth and metastasis, making them potential therapeutic targets.

[0003] Mesenchymal stromal cells (MSCs) are generally considered to be a non-epithelial, non-hematopoietic cellular component that plays a crucial role in tissue remodeling, inflammatory responses, epithelial cell growth, and immunosuppression. MSCs lack typical lineage markers and generally express vimentin, collagen, platelet-derived growth factor receptor α / β, and mucin, with variable distribution patterns across tissues and cell types. Recent advances in single-cell transcriptomics have enabled systematic analysis of cell populations to dissect colorectal diseases, including inflammatory bowel disease and colorectal cancer, at unprecedented resolution. Distinct stromal cell populations may play distinct roles in the development of IBD. For example, one stromal cell population has been reported to reside in the crypt niche and carry out normal repair and regenerative responses, while another stromal cell population has proinflammatory characteristics that contribute to disease severity. Furthermore, inflammatory fibroblasts expressing IL13RA2 and IL11 have been associated with resistance to anti-TNF therapy in IBD patients. Stromal cell heterogeneity has also been implicated in the outcome of CRC progression. Myofibroblasts and cancer-associated fibroblasts have been reported to be preferentially enriched in CRC tumors. In addition, myofibroblast-related gene signatures were also identified as one of the main features of CRC shared molecular subtype 4. This subtype presents the characteristics of tumor stroma enrichment and TGF-β signaling-mediated extracellular matrix remodeling. However, since the stromal cells used in the above studies only accounted for a small part of the total sequenced cell population, this hindered their in-depth study at high resolution. There is an urgent need to understand the definitive functions of stromal subtypes, especially regarding their interactions with other cells in the tumor microenvironment (TME). In this regard, single-cell RNA sequencing (scRNA-seq) has revealed the importance of cell interactions in a variety of cancers.

[0004] Recently, it has been reported that low infiltrating immune cells in the TME are associated with poor prognosis in CRC patients, and that tumor-associated macrophages (TAMs) localized at the edge of the tumor can prevent cytotoxic lymphocytes (CTLs) from infiltrating into the tumor core. The two types of TAMs have different inflammatory and angiogenic characteristics and respond oppositely to CSF1R blockade therapy. Macrophages expressing M2 macrophage markers (e.g., CD163, DC-SIGN) or fibroblasts expressing FSP1, FAP are positively correlated with poor prognosis in CRC patients. A TAM subtype with unique characteristics, called SPP1, has recently been reported. +Macrophages, which have immunosuppressive properties and are positively correlated with EMT markers, may serve as potential targets for combating tumor growth and metastasis. However, the interactions between myeloid cells and other cell types in the CRC tumor microenvironment remain to be further investigated. Summary of the Invention

[0005] Currently, most scRNA-seq studies on CRC are based on a few specific cell types, which affects the comprehensive understanding of cell types and possible cell interactions in the tumor microenvironment. The cost of scRNA-seq is relatively high, and the number of patients enrolled in existing single-cell sequencing studies is limited. Therefore, public big data is used to find cell subpopulations and tumor progression and related properties to more comprehensively understand the relationship between each subpopulation and disease progression. This paper will perform single-cell sequencing analysis based on 5 CRC patients, describe the clustering characteristics, and the correlation between each subpopulation in the public database, and further explore the interaction regulatory network and spatial localization between highly correlated cell types. The spatial transcriptome is further used to describe the characteristics of cell subpopulations specific to the tumor boundary.

[0006] The present invention proposes a colorectal cancer tumor border cell subpopulation biomarker and a colorectal cancer detection biomarker based on any one or a combination of several of Chemerin, FAP, and SPP1.

[0007] The present invention also proposes the use of any one or a combination of Chemerin, FAP, and SPP1 as colorectal cancer tumor boundary and colorectal cancer biomarkers. The application uses any one or a combination of Chemerin, FAP, and SPP1 as targets to prepare diagnostic reagents for colorectal cancer tumor boundary definition and colorectal cancer occurrence and / or metastasis, or to prepare drugs for colorectal cancer treatment, or to prepare reagents for predicting colorectal cancer survival time, or to prepare reagents for colorectal cancer prognosis evaluation.

[0008] The application is to immunize animals with any one or a combination of Chemerin, FAP, SPP1 to prepare antibodies.

[0009] Alternatively, the application is to prepare immune cells, proteins and / or small molecules targeting any one or a combination of Chemerin, FAP, and SPP1.

[0010] The present invention also provides a method for diagnosing colorectal cancer in a subject, or a method for preventing or treating colorectal cancer in a subject in need, or a method for predicting or prognosticating the survival time of a patient with colorectal cancer.

[0011] The method comprises: determining the level of any one or a combination of several of Chemerin, FAP, and SPP1 in a sample obtained from the patient or subject, wherein the level is used to diagnose whether the subject has colorectal cancer, or to predict or evaluate the patient's survival time; or

[0012] The method comprises: using any one or a combination of several of the Chemerin, FAP, and SPP1 as a target to prevent or treat colorectal cancer in the patient.

[0013] In the use or method described in the present invention, the sample for measuring the biomarker is a sample of tumor tissue or blood obtained from the patient, or a sample of intestinal tissue or blood from the subject.

[0014] In the uses or methods of the present invention, the level of any one or more of Chemerin, FAP, and SPP1 is determined at the protein level; or, the level of any one or more of Chemerin, FAP, and SPP1 is determined at the nucleic acid level.

[0015] In the uses or methods of the present invention, when the level of any one or more of Chemerin, FAP, and SPP1 is determined at the protein level, the level of any one or more of Chemerin, FAP, and SPP1 is determined by immunohistochemistry;

[0016] When the level of any one or more of Chemerin, FAP, and SPP1 is determined at the nucleic acid level, the level of any one or more of Chemerin, FAP, and SPP1 is determined by quantifying the mRNA encoding any one or more of Chemerin, FAP, and SPP1.

[0017] In the use or method described in the present invention, the lower the level of any one or more of Chemerin, FAP, and SPP1, the higher the probability that the patient will have a long survival time.

[0018] In the application or method described in the present invention,

[0019] The diagnosis comprises the following steps: i) determining the level of Chemerin in a sample obtained from the subject; ii) comparing the level determined in step i) with a predetermined reference value; iii) when the level determined in step i) is higher than the predetermined reference value, the individual being measured is a colorectal cancer patient, or when the level determined in step i) is close to the predetermined reference value, the individual being measured is a healthy individual; or,

[0020] The diagnosis comprises the following steps: i) determining the levels of Chemerin, FAP, and SPP1 in an intestinal sample obtained from the subject; ii) comparing the levels determined in step i) with predetermined reference values; iii) when the intestinal SPP1 determined in step i) is + Macrophages or FAP + Characteristic upregulation of any one or more of the above on the surface of fibroblasts, wherein the individual being tested is a colorectal cancer patient; or

[0021] The prediction or prognostic assessment comprises the following steps: i) determining the level of Chemerin in a sample obtained from the patient; ii) comparing the level determined in step i) with a predetermined reference value; iii) when the level determined in step i) is lower than the predetermined reference value, the patient has a good prognosis, or when the level determined in step i) is higher than the predetermined reference value, the patient has a poor prognosis; or,

[0022] The prevention or treatment comprises the following steps: preventing or treating colorectal cancer in a subject in need thereof by targeting colorectal cancer tumor-specific biomarkers and cells, comprising: administering to the subject a pharmaceutical composition comprising a pharmaceutically acceptable carrier, an effective amount of a regulator of colorectal cancer tumor-specific biomarkers or cells, and optionally another therapeutic agent, thereby preventing or treating colorectal cancer; wherein the colorectal cancer tumor-specific biomarker is selected from Chemerin, FAP and SPP1, and the cells are SPP1 + Macrophages and FAP + Fibroblasts.

[0023] In the application or method of the present invention, the regulator is selected from small molecule chemical agents, antisense oligonucleotides, small interfering RNA (siRNA), short hairpin RNA (shRNA), therapeutic vaccines, antibodies, and biologically active fragments or homologs of the antibodies.

[0024] The present invention also proposes a biomarker Chemerin, FAP, SPP1 regulator, which includes small molecule compounds, antisense oligonucleotides, small interfering RNA (siRNA), short hairpin RNA (shRNA), polypeptide and protein drugs, etc.

[0025] The present invention also provides an antibody, which includes an activating or inhibitory antibody against the biomarkers Chemerin, FAP, and SPP1, and the antibody has specific binding ability to the biomarkers Chemerin, FAP, and SPP1.

[0026] The present invention also proposes a therapeutic vaccine, which includes polypeptide fragments or full-length proteins of the biomarkers Chemerin, FAP, and SPP1 proteins. The therapeutic vaccine can stimulate a specific immune response.

[0027] The present invention also provides a drug / drug composition comprising a regulator of the biomarkers Chemerin, FAP, and SPP1, a therapeutic vaccine, and / or an antibody against Chemerin, FAP, and SPP1 as described above.

[0028] The present invention also proposes a detection reagent / kit, which comprises any one of Chemerin, FAP and SPP1 or different combinations thereof, and is a diagnostic kit for early diagnosis, concomitant diagnosis and prognosis evaluation of colorectal cancer based on protein markers and nucleic acid sequences.

[0029] The present invention also proposes the use of the above-mentioned biomarkers Chemerin, FAP, SPP1 regulator, or the antibody, or the therapeutic vaccine, or the drug / drug composition, or the detection reagent / kit in the preparation of diagnostic reagents for the occurrence and / or metastasis of colorectal cancer, or the preparation of drugs for the treatment of colorectal cancer, or the preparation of reagents for predicting the survival time of colorectal cancer, or the preparation of reagents for evaluating the prognosis of colorectal cancer.

[0030] The present invention performed deconvolution calculations on 14 colorectal cancer datasets and found that the infiltration ratios of the two cell populations showed a high correlation. Cell interaction analysis revealed that the two cell populations may play a regulatory role through the interaction of RARRES2 / CMKLR1. It also revealed that the Chemerin protein encoded by RARRES2 showed a higher expression level in colorectal cancer plasma compared with healthy human plasma, suggesting that it can be used as a diagnostic marker for colorectal cancer. In addition, SPP1 + Macrophages can regulate FAP through genes encoding TGFB1, IL1A, and IL1B + The ability of fibroblasts to produce extracellular matrix promotes the formation of tumor fibrosis "wall"-like structures, thereby creating an immune "rejection" microenvironment. + Fibroblasts or SPP1 + The progression-free survival of colorectal cancer patients with macrophages is significantly lower than that of the control group. Given that this subpopulation specifically expresses FAP and SPP1, it is further disclosed that patients with high expression of these two proteins in the bladder cancer data set also have a lower response rate to PD-L1 treatment. The present invention also proposes FAP + Fibroblasts and SPP1 +Macrophages are also spatially localized in colorectal cancer, forming a "wall"-like tumor boundary around the tumor center, thereby preventing immune cells from entering the tumor center. These cellular characteristics and cell interaction patterns can be used to effectively diagnose, prevent, treat, and assess prognosis and survival for CRC. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Classification of myeloid cells and mesenchymal stromal cells in colorectal cancer; (A) UMAP cluster distribution map analyzes the classification of myeloid cells in colon cancer and adjacent tissues. (B) SPP1 + The proportion of macrophages in adjacent tissues and colorectal cancer tissues. (C) UMAP cluster distribution diagram analyzes the classification of mesenchymal stromal cells in colon cancer and adjacent tissues. (D) FAP + The proportion of fibroblasts in adjacent cancer tissues and colorectal cancer tissues.

[0032] Figure 2 For FAP + Fibroblasts and SPP1 + The Spearman correlation analysis of macrophages in each data set showed that the p values ​​were all less than 0.001 and the Rs were all greater than 0.3.

[0033] Figure 3 For FAP + Fibroblasts and SPP1 + Macrophage infiltration in different stages and its relationship with patient survival.

[0034] (A)FAP + Fibroblasts and SPP1 + Comparison of the infiltration ratio of macrophages in different stages of colorectal cancer, from left to right: Stage I, Stage II, Stage III, Stage IV. (B) FAP in COAD colorectal cancer samples + Fibroblasts and SPP1 + Kaplan-Meier survival analysis of the relationship between the proportion of macrophage tumor infiltration and patient survival.

[0035] Figure 4 For FAP + Fibroblasts and SPP1 + Kaplan-Meier analysis of the relationship between different macrophage infiltration ratios and patient survival.

[0036] Figure 5 Immunofluorescence shows the spatial localization of FAP+ and SPP+ cells.

[0037] (A) From left to right, EPCAM (epithelial), DAPI (nucleus), SPP1, and FAP. The figure shows three colorectal cancer tissues stained with immunofluorescence. (B) Adjacent and non-adjacent FAP in each field of view + SPP1 + Statistical analysis was performed on the cell percentage. **p<0.05

[0038] Figure 6 Uncovering FAPs for spatial transcriptomes + Fibroblasts and SPP1 + Colocalization was observed in macrophages.

[0039] (A) According to FAP + Fibroblasts and SPP1 + Macrophage gene signatures were used to score cell clustering at the spatial transcriptional level in four colorectal cancer samples.

[0040] (B) Four colorectal cancer idle samples with FAP + Fibroblasts or SPP1 + Functional enrichment of differentially expressed genes obtained from the macrophage feature array.

[0041] (C) FAP in colorectal cancer idle samples + Fibroblasts, SPP1 + Spatial distribution of macrophage and epithelial cell scores: higher expression indicates darker color. Spatial distribution of CD3D, CD8A, and CD4 gene expression.

[0042] Figure 7 For FAP + Fibroblasts and SPP1 + Analysis of macrophage regulatory networks.

[0043] (A) Left: FAP + Fibroblasts regulate SPP1 + The relative expression values ​​of macrophage ligands in 8 fibroblasts, the bottom graph is and FAP + Relative expression of ligand-matched receptors for fibroblasts in seven myeloid cell subsets. The middle figure is FAP from scRNA-seq. + Fibroblasts and SPP1 +Heatmap of important ligand-receptor pairs between macrophages, calculated by NichenetR. The X-axis is the receptor calculated by NichenetR, and the Y-axis is the ligand. (B) UMAP cluster distribution diagram showing the relative expression of RARRES2 in various mesenchymal stromal cell subsets. (C) UMAP cluster distribution diagram showing the relative expression of CMKLR1 in various myeloid cell subsets. (D) SPP1 calculated by NichenetR + Macrophage regulation of FAP + The top-ranked ligand activity in fibroblasts. (E) Distribution of the relative expression levels of each ligand gene in each myeloid cell subset and the proportion of expressing cells. (F) SPP1 in scRNA-seq calculated by NichenetR + Macrophages and FAP + Heat map of important ligand-receptor pairs between fibroblasts. (G) Each SPP1 + Ligands on macrophages act on FAP + Heat map of the relative likelihood of each target gene being regulated in fibroblasts. (H) SPP1 + Ligands on macrophages act on FAP + Pathway analysis of target genes enriched in fibroblasts. (I) Graph showing SPP1 + Ligands on macrophages regulate FAP + Fibroblasts belong to the target genes and potential signaling pathways of the extracellular matrix and the corresponding first three ligands.

[0044] Figure 8 To compare plasma chemerin levels in colorectal cancer patients and healthy controls.

[0045] Chemerin levels in healthy volunteers (n=17) and CRC patients (n=20) were measured using ELISA. Results are presented as mean ± SD, unpaired t-test, *p < 0.05.

[0046] Figure 9 For FAP + Fibroblasts and SPP1 + Correlation between macrophage infiltration and immunotherapy.

[0047] (A) Different FAPs in the COAD dataset of the TCGA database + The proportion of lymphocyte infiltration in colorectal cancer samples with fibroblast and SPP1+ macrophage infiltration. (B) Kaplan-Meier analysis shows the relationship between FAP expression in the IMvigor210 dataset and patient survival. (C) Kaplan-Meier analysis shows the relationship between SPP1 expression in the IMvigor210 dataset and patient survival. (D)

[0048] Kaplan-Meier analysis shows the relationship between the combined expression of FAP and SPP1 in the IMvigor210 dataset and patient survival. (E) The relationship between the combined expression of FAP and SPP1 and the patient response rate to PD-L1. CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease. DETAILED DESCRIPTION

[0049] The present invention is further described in detail with reference to the following specific examples and accompanying drawings. The processes, conditions, experimental methods, etc. for implementing the present invention, except for those specifically mentioned below, are common knowledge and common common sense in the art and are not particularly limited by the present invention.

[0050] Example 1 Immune microenvironment cell map of adjacent cancer tissues and cancer tissues

[0051] 1. Materials and Methods

[0052] 1.1 Single cell isolation from human intestinal tissue

[0053] (1) Freshly resected human intestinal cancer tumor tissue and adjacent tissue samples were placed in a 50 mL centrifuge tube containing RPMI 1640 complete medium containing 10% FBS, placed on ice, and sent to the laboratory as soon as possible.

[0054] (2) Rinse with 10 mL of PBS to remove blood and fat from the sample.

[0055] (3) A portion of the tissue treated in step (2) was fixed with 1% PFA, and the remaining tissue was weighed and subsequently prepared for cell suspension.

[0056] (4) The mucosal layer of the adjacent tissue was removed and cut into pieces approximately the size of soybeans using scissors. The pieces were placed in 10 mL of epithelial removal solution (PBS containing 5 mM EDTA, 15 mM HEPES, 10% FBS, and 1 mM DTT) and placed in a 50 mL centrifuge tube. The tubes were shaken in a bacterial shaker at 200 rpm and 37°C for 1 h.

[0057] (5) Human intestinal cancer tumor tissue was also cut into small pieces and placed in a PBS solution containing 65 mM DTT. The pieces were then shaken in a bacterial shaker at 200 rpm and 37°C for 15 minutes.

[0058] (6) After completing the above steps, shake vigorously manually for 2 minutes.

[0059] (7) Wash the tissue twice with 10 mL of PBS after shaking in steps (4) and (5) to remove residual EDTA, DTT, etc.

[0060] (8) Cut the human intestinal cancer tumor tissue and adjacent tissue into 3mm 2 The pellets were cut into pieces of 100 μg / mL and digested for 1 h in complete RPMI 1640 medium containing 0.38 μg / mL collagenase VIII and 0.1 mg / mL DNase I. The instrument conditions were 200 rpm and 37°C.

[0061] (9) After digestion, shake vigorously manually for 5 minutes and repeatedly pipette with a 10 mL syringe.

[0062] (10) Add 20 mL of PBS and filter the digested tissue using a 100 μm cell strainer.

[0063] (11) Centrifuge at 1800 rpm for 5 minutes.

[0064] (12) Discard the supernatant and resuspend in RPMI 1640 containing 10% FBS.

[0065] (13) For whole-tissue single-cell sequencing or interstitial cell staining, the samples are washed again and used for subsequent single-cell transcriptome sequencing.

[0066] 1.2 Single-cell RNA sequencing

[0067] (1) Freshly prepared paracancerous or cancerous cells in step 1.1 above; or sorted ILCs cells from various tissues.

[0068] (2) Test the viability of the above cells and ensure that it is above 85%.

[0069] (3) Resuspend the cells according to the cell concentration recommended by 10X Genomics' Chromium Single Cell 3′ Reagent Kits v3, and then generate single-cell GEMs, reverse transcription and purification, cDNA amplification and cDNA purification, quality inspection and quantification.

[0070] (4) Subsequently, the cDNA is fragmented and undergoes steps such as adapter ligation, library amplification, and quality control.

[0071] (5) Sequencing was completed using the Illumina platform (Novaseq 6000) in PE150 mode.

[0072] (6) After sequencing, the original BCL format file is obtained for subsequent analysis.

[0073] 1.3 Sequencing data alignment, quality control, and normalization

[0074] (1) Raw sequencing reads were quality controlled using FastQC software v0.11.9 (https: / / www.bioinformatics.babraham.ac.uk / projects / fastqc / ).

[0075] (2) The sequencing data were converted from the bcl format to the FASTQ format using Illumina bcl2fastq2 conversion software v2.20 (https: / / support.illumina.com / downloads / bcl2fastq-conversion-software-v2-20.html).

[0076] (3) The cells were then processed, aligned, and counted using the Cell Ranger single-cell software suite v.2.2 (https: / / support.10xgenomics.com / single-cell-gene-expression / software / pipelines / latest / ) according to the software standard workflow and default parameters.

[0077] (4) Briefly, the present invention uses a standard cell sequencing workflow to align FASTQ reads with the human genome GRch38 genome.

[0078] (5) Then, some sequencing reads are filtered out based on the quality score of base recognition, and corresponding cell labels and UMIs are assigned to each read.

[0079] (6) During quality control analysis using Seurat, the UMI matrix was filtered to remove genes expressed in fewer than three cells, cells with fewer than 200 genes, cells with more than 4,000 genes, and cells with a high percentage of mitochondrial genes (more than 8%). The resulting matrix was then normalized, transformed using a “scaling factor” (default 10,000), and log transformed using the “LogNormalize” function in Seurat for downstream analysis.

[0080] 1.4 Cell classification, filtration of contaminated cells, and batch effect correction

[0081] For both adjacent and cancerous tissues, we classified the major clusters based on known cell type markers. After classification, we subclassified the clusters and re-clustered them. Finally, we reduced the dimensionality and removed other cell types. Finally, we labeled all pure cells and merged them into a unified cluster, correcting for batch effects. We used the "RunHarmony" method to correct for batch effects in clustering all cells in adjacent and cancerous tissues.

[0082] 1.5 UMAP Dimensionality Reduction Analysis

[0083] For single-cell sequencing of all adjacent and cancerous tissues, Seurat's "FindVariableGenes" tool was used to identify the top 2,000-3,000 variable genes, based on cell type, for subsequent dimensionality reduction clustering. Different PCs were assigned to different subpopulations for subsequent "FindClusters" and "RunUMAP" analyses.

[0084] 1.6 Differential gene analysis

[0085] Differentially expressed genes in each subgroup were analyzed using the "FindAllMarkers" function in the Seurat package. The nonparametric Wilcoxon rank sum test was used with Bonferroni correction to obtain p-values ​​for all genes in the dataset, as well as adjusted p-values. The logarithmic fold change (logFC) of expression values ​​was calculated using the following parameters, and p-values ​​for all variable genes in each cluster were obtained: min.pct = 0.05, min.diff.pct = 0.1, logfc.threshold = 0.25. Genes were log-transformed and scaled for expression to generate heatmaps.

[0086] 2. Experimental results

[0087] Using the transcriptome data of 54,103 cells obtained by 10X single-cell sequencing, Harmony was used to correct the batch effects of different samples, and UMAP was used to reduce the dimensionality of the data. According to the surface markers of the cell populations, they were divided into three major groups: epithelial cells (EPCAM), stromal cells (COL1A1, COL3A1) and immune cells (PTPRC, CD19, MZB1, CD3E, CD14). Cell annotation was performed for each cell population based on known cell markers. Finally, 58 cell subsets were obtained, including 10 epithelial cell populations, 10 fibroblast subsets, 5 endothelial cell subsets, 10 NK, ILCs, T cell subsets, 9 myeloid cell subsets, 2 mast cell subsets, 7 B cell subsets and 2 plasma cell subsets. The specific marker genes of each cell subset were further determined.

[0088] Furthermore, myeloid cells can be divided into activated DCs (labeled as activated DCs), cDC1, cDC2, macrophages expressing the macrophage marker C1QC but not MRC1 (labeled as C1QC + MRC1 - macrophages), macrophages that specifically express SPP1 (labeled as SPP1 + macrophages), THBS1 + Macrophages (labeled as THBS1 + macrophages), monocytes (VCAN + monocytes), neutrophils and proliferating myeloid cells ( Figure 1 A) To further explore the cell subpopulations closely related to tumors, the present invention statistically analyzed the proportion of each subpopulation in adjacent and cancerous tissues. The results showed that SPP1 + The proportion of macrophages in tumors increased significantly, from almost undetectable to approximately 10% of the total myeloid cells ( Figure 1 B).

[0089] Stromal cells were classified into CD24 + Fibroblasts (CD24 + fibroblasts), NT5E + Fibroblasts (NT5E + fibroblasts), DES + Myofibroblasts (DES + myofibroblasts), FAP + Fibroblasts (FAPs + fibroblasts), FGFR2 + Fibroblasts (FGFR2 + fibroblasts), ICAM1 - Trojan cells (ICAM1 - telocytes), ICAM1 + Trojan cells (ICAM1 + telocytes), MFAP5 + Myofibroblasts (MFAP5 + myofibroblasts), pericytes and proliferating fibroblasts ( Figure 1C) Among them, FAP + The proportion of fibroblasts in tumors is significantly higher than that in adjacent tissues ( Figure 1 D).

[0090] Example 2 Analysis of multiple colorectal cancer databases reveals FAP + Fibroblasts and SPP1 + Positive correlation with macrophage infiltration

[0091] 1. Materials and Methods

[0092] Using the CibersortX website, we projected the transcriptome lists for each cell subpopulation from the single-cell transcriptome analysis onto multiple colorectal cancer RNAseq and microarray expression datasets, using a permutation test parameter of 500. Quantile normalization was used for the microarray gene expression dataset, while the RNAseq data were not normalized. The output data was a matrix of the infiltration percentages of each cell type in each sample. Subsequently, we analyzed the correlations between the infiltration percentages of each cell population and different clinical characteristics in R.

[0093] 2. Experimental results

[0094] In order to explore the correlation of infiltration of each cell subpopulation in the tumor microenvironment, the present invention uses CibersortX to map the characteristic genes of each subpopulation obtained by the single cell transcriptome of the present invention to the colon cancer bulk RNA and chip (microarray) data, and split each cell type to obtain the cell infiltration ratio. The colorectal cancer data sets used include TCGA's COAD and rectal cancer (Rectum Adenocarcinoma, READ), and GEO's 12 colorectal cancer expression matrices, namely GSE39582, GSE17536, GSE17537, GSE23878, GSE33113, GSE41568, GSE37892, GSE20916, GSE21510, GSE18105, GSE13294 and GSE14333. By calculating the correlation of the infiltration ratios of these 58 cell subpopulations, it was found that FAP + fibroblasts and SPP1 + The infiltration of macrophages in all colorectal cancers is positively correlated ( Figure 2 ).

[0095] Example 3 FAP + Fibroblasts and SPP1 + Macrophages negatively correlated with patient survival

[0096] 1. Materials and Methods

[0097] Survival analysis

[0098] Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated using the Cox proportional hazards model, and Kaplan-Meier survival curves were modeled using the survfit function. The "maxstat.test" function from the R package maxstat was used to repeatedly test all potential cut points to find the maximum log-rank statistic for dichotomizing cell population infiltration or gene expression, and then to select the maximum log-rank statistic. Kaplan-Meier survival curves were compared using the two-sided long-rank test.

[0099] 2. Experimental results

[0100] Furthermore, by detecting FAP + Fibroblasts and SPP1 + The correlation between macrophage infiltration and colorectal cancer stage was found to be lower in stage I colorectal cancer, while the infiltration ratios in other stages of colorectal cancer were higher than those in stage I. Figure 3 A). The relationship between its infiltration ratio and the survival of colorectal cancer patients deserves further study. Therefore, the present invention statistically analyzed the relationship between the patient's progression-free survival (PFS) and these two types of cell infiltration. Kaplan-Meier survival analysis found that in COAD colorectal cancer samples, the tumor had high infiltration of FAP. + The survival period of cancer patients with fibroblasts is significantly shorter than that of patients with low infiltration; similarly, high infiltration SPP1 in tumors + Cancer patients with macrophages have shorter survival periods than those with low-invasive disease ( Figure 3 B) Subsequently, the present invention grouped the COAD data according to the ratio of these two types of cells infiltrating the patient's tumor and found that FAP + fibroblasts and SPP1 + Patients with high levels of macrophages had the lowest survival time, followed by FAP. + fibroblasts high, SPP1 + Patients with low macrophages, and then FAP + Patients with low fibroblasts ( Figure 4 ). FAP + The infiltration ratio of fibroblasts may be a key factor affecting patient survival, and SPP1 + The co-existence of macrophages exacerbates the poor survival of patients.

[0101] Example 4 FAP+ Fibroblasts and SPP1 + Macrophages spatially colocalize

[0102] 1. Materials and Methods

[0103] 1.1 Immunofluorescence detection of colocalization of interstitial cells and macrophages

[0104] (1) Part of the tumor tissue was excised and placed in a fixation buffer (PBS containing 1% PFA) and fixed at 4°C overnight.

[0105] (2) Transfer the sample to PBS containing 30% sucrose and incubate at 4°C overnight.

[0106] (3) Place the sample in OCT and freeze at -80°C.

[0107] (4) Cut the tissue into 10 μm slices.

[0108] (5) After sectioning, rehydrate the sections in PBS for 10 min.

[0109] (6) Wash the sections with PBS and then soak them in pre-cooled methanol at -20°C for 30 minutes.

[0110] (7) Wash the sections three times with PBS and draw a circle around the sample on the sections with an immunohistochemical pen. Then add blocking buffer for 1 hour at room temperature (in a humidified chamber). (Blocking buffer: 0.3% Triton X-100, 1% BSA, 1% FBS, and 0.1 mol / L Tris-HCl buffer, with goat serum added at a ratio of 1:100).

[0111] (8) Rabbit anti-human FAP antibody (1:150) was added to the sections and incubated at room temperature for 3 hours. After that, the sections were washed three times with PBS.

[0112] (9) Use AF647-labeled goat anti-rabbit, PE-labeled mouse anti-human SPP1, and Alexa Fluor 488-labeled mouse anti-human EPCAM antibodies at dilution ratios of 1:200, 1:60, and 1:200, respectively, and incubate in a humidified chamber at 4°C overnight.

[0113] (10) Wash three times with PBS.

[0114] (11) The sections were sealed with anti-fade sealing medium containing DAPI.

[0115] (12) Cover with a coverslip and observe the staining effect under a Confocal microscope.

[0116] 1.2 Spatial transcriptome analysis

[0117] Spatial transcriptomics slides were obtained from capture regions of four CRC patients. Gene expression information from ST slides was captured using the 10x Genomics Visium Spatial platform using a standard workflow using spatially barcoded mRNA-binding oligonucleotides. Raw sequencing reads for spatial transcriptomics were quality-checked and mapped using Space Ranger v1.1. Gene locus matrices generated after data processing of spatial transcriptomics samples were analyzed using the Seurat package (version 3.2.1) in R. Loci expressing fewer than 200 genes, genes with fewer than 10 counts, or genes expressed in fewer than three loci were filtered out. Across-locus normalization was performed using the LogVMR function. Dimensionality reduction and clustering of the top 30 PCs were performed using independent principal component analysis at a resolution of 1.1. Gene scoring for scRNA-seq or ST was performed in Seurat using the AddModuleScore function with default parameters. Spatial feature expression plots were generated using the SpatialFeaturePlot function in Seurat (version 3.2.1).

[0118] 2. Experimental results

[0119] Given FAP + Fibroblasts and SPP1 + The positive correlation between macrophage infiltration and colorectal cancer tissue in the present invention is speculated to exist spatial interaction. The present invention used immunofluorescence to stain SPP1 + Macrophage-specific markers SPP1 and FAP + FAP, a marker protein for fibroblasts, was found to be closely located in colorectal cancer tissues. Figure 5 A), and quantitative analysis revealed that about 70% of SPP1 + Cells and FAPs + Cells are located closer ( Figure 5 B).

[0120] To further evaluate FAP + Fibroblasts and SPP1 + The spatial localization relationship of macrophages. The present invention performed spatial transcriptomic sequencing (ST) on tumor tissue sections from four CRC patients. The spatial resolution of the 10x Genomics Visium platform generally accommodates about 10 cells per spot. Based on this, cell types located at the same spot have a proximity relationship at the spatial level. Through unsupervised clustering methods, it can be found that FAP is present in the ST results of the four colorectal cancer tissues.+ Fibroblasts / SPP1 + The clustering characteristics of macrophages suggest that they are adjacent cells in space ( Figure 6 A) FAP + Fibroblasts / SPP1 + The macrophage clustering features showed the characteristics of promoting the "wall"-like structure of fibroproliferative structures, such as extracellular matrix formation, collagen fibril organization and response to TGF-β ( Figure 6 B). In addition, these cell clusters may exhibit characteristics of surrounding epithelial cells and excluding immune cells (such as T and B cells) from entering the tumor core ( Figure 6 C).

[0121] Example 5 FAP + Fibroblasts and SPP1 + Macrophages can regulate each other to remodel the extracellular matrix

[0122] 1. Materials and Methods

[0123] 1.1 NichenetR analysis

[0124] The R package NichenetR comes with "ligand_target_matrix.rd", "lr_network.rds" and "weighted_networks.rds" as the reference data set of ligand receptor regulatory network. The present invention selects genes with expression ratio greater than 0.1, selects the top 100 ligands, the top 1000 targets, and up-regulated genes for analysis. + fibroblasts to SPP1 + When macrophages are affected, select THBS1 + macrophages as a reference dataset. + Macrophages for FAP + When fibroblasts act, they select FGFR2 + fibroblasts and ICAM1 + We used telocytes as a reference dataset and used the "nichenet_seuratobj_cluster_de" function in NichenetR to find potential interacting receptor ligands. We also used the "ligand_receptor_heatmap_bonafide" function to output a heatmap.

[0125] 2. Experimental results

[0126] To further analyze FAP+ Fibroblasts and SPP1 + The present invention uses NichenetR to analyze the ligand-receptor interaction network between these two groups of cells.

[0127] Using FAP + fibroblasts as ligand cells, SPP1 + NichenetR analysis was performed on macrophages as receptor cells, and the relative expression of these ligand receptors in these two types of cells was calculated. + High expression levels in fibroblasts, receptors in SPP1 + The pathways with relatively high expression levels in macrophages may be ligand-receptor pairs that play an important role. The results showed that COL1A1 / ITGB1, LAMA1 / ITGB1, INHBA / ACVR1, INHBA / ACVRL1, CCL3 / CCR5, CCL3 / CCR1, RARRES2 / CMKLR1, TGFB1 / ACVR1 and WTN5A / FZD2 had relatively high ligand and receptor scores and high expression levels in the corresponding cells ( Figure 7 A). The present invention examined the expression level of RARRES2 in mesenchymal stromal cells and found that FAP + fibroblasts do have relatively high expression ( Figure 7 B); CMKLR1 is in SPP1 + Macrophages have relatively high expression ( Figure 7 C).

[0128] Fibroblasts are the main producers of extracellular components, including growth factors, cytokines and extracellular matrix components, which may contribute to the formation of fibroproliferative structures. Based on this, the present invention studied SPP1 + Do macrophages promote FAP? + ECM remodeling ability of fibroblasts. SPP1 + Macrophages showed higher TGFB1, IL1B, and IL1A ligand activities and relatively high gene expression ( Figure 7 D) In ​​addition, TGFB1 and FAP + TGFBR3, ACVRL1 and TGFBR1 expressed on fibroblasts bind to each other, while IL1B / IL1A bind to FAP + Interaction of IL1R1 or IL1RAP on the fibroblast cell membrane ( Figure 7E), leading to the expression of target genes encoding collagen or matrix metallopeptidase in these cells ( Figure 7 F). These target genes are important components of the desmoplastic response, and 35 of the 100 predicted targets encode proteins involved in ECM remodeling, including extracellular matrix components (e.g., collagens [COL10A1, COL11A1, COL1A1, COL1A2, COL3A1, COL5A1, COL8A1], fibronectin [FN1], and integrins [ITGA5, ITGB5]), remodeling proteins (e.g., lysyl oxidase family [LOX, LOXL1, LOXL2]), and matrix metalloproteinases (ADAM17, MMP1, MMP14, MMP2, MMP3, TIMP1, TIMP2, TIMP3) ( Figure 7 F). Further KEGG pathway enrichment of the predicted genes revealed that these genes were mainly enriched in cytokine-cytokine receptor interaction, extracellular matrix pathway, TNF signaling pathway and TGF-β signaling pathway ( Figure 7 G). Signaling pathway analysis of the top three ligands and ECM-related genes revealed 40 downstream signaling pathways that may be connected to SPP1. + Ligands secreted by macrophages and promoting FAP + Fibroblasts express ECM-related targets, thereby promoting the generation of "wall"-like structures.

[0129] Example 6 Chemerin levels in CRC patients' plasma are higher than those in healthy controls

[0130] 1. Materials and Methods

[0131] 1.1 Detection of plasma chemerin by enzyme-linked immunosorbent assay

[0132] (1) Take a 96-well plate and add 100 μL of Capture Antibody (stock concentration: 720 μg / mL, working concentration: 4 μg / mL) to each well.

[0133] (2) Cover with plastic wrap and store overnight at room temperature.

[0134] (3) Add 200 μL Wash Buffer (PBS containing 0.05% Tween 20, pH 7.2-7.4) to each well, let it stand for 1 min, discard the liquid and tap dry, and repeat 3 times.

[0135] (4) Add 300 μL Reagent Diluent (PBS containing 1% BSA, pH 7.2-7.4, 0.2 μL filtered) to each well and block at room temperature for 1 hour.

[0136] (5) Discard the liquid and tap dry. Add 200 μL Wash Buffer (PBS containing 0.05% Tween 20, pH 7.2-7.4) to each well, let it stand for 1 min, discard the liquid and tap dry. Repeat twice.

[0137] (6) In a 96-well deep-well plate, use Reagent Diluent to prepare samples of the required concentration for the experiment.

[0138] (7) Prepare standards using Eppendorf tubes and perform serial dilutions. The concentrations of the standards are 2000 pg / mL, 1000 pg / mL, 500 pg / mL, 250 pg / mL, 125 pg / mL, 62.5 pg / mL, and 31.25 pg / mL.

[0139] (8) In a 96-well plate, add 100 μL of sample or standard dissolved in Reagent Diluent to each well (duplicate wells for each sample), protect from light, and incubate at room temperature for 2 h.

[0140] (9) Discard the liquid and tap dry. Add 200 μL Wash Buffer (PBS containing 0.05% Tween 20, pH 7.2-7.4) to each well, let it stand for 1 min, discard the liquid and tap dry. Repeat 4 times.

[0141] (10) Add 100 μL of detection antibody dissolved in Reagent Diluent (stock concentration 36 μg / mL, working concentration 200 ng / mL) to each well, protect from light, and incubate at room temperature for 2 h.

[0142] (11) Discard the liquid and tap dry. Add 200 μL Wash Buffer (PBS containing 0.05% Tween 20, pH 7.2-7.4) to each well, let it stand for 1 min, discard the liquid and tap dry. Repeat 4 times.

[0143] (12) Add 100 μL of Streptavidin-HRP (1:200) dissolved in Reagent dilution to each well and incubate at room temperature for 20 min in the dark.

[0144] (13) Discard the liquid and tap dry. Add 200 μL Wash Buffer (PBS containing 0.05% Tween 20, pH 7.2-7.4) to each well, let it stand for 1 min, discard the liquid and tap dry. Repeat 5 times.

[0145] (14) Add 100 μL of TMB to each well and incubate at room temperature for 20 min in the dark.

[0146] (15) Add 50 μL Stop Solution (2N H2SO4) to each well and tap gently to mix thoroughly.

[0147] (16) The absorbance at 450 nm and 570 nm was measured using an enzyme-labeled instrument, and the concentration of each well was calculated using a four-parameter regression method based on the standard curve.

[0148] 2. Experimental results

[0149] Chemerin is a secretory protein. It has been reported that its content may be correlated with the survival of patients with right colon cancer. Can it be used as a potential molecular marker for CRC patients? To answer this question, the present invention collected plasma from healthy subjects and CRC patients and tested whether there was a difference in chemerin in the plasma. The results showed that compared with the plasma of healthy subjects, the plasma of colorectal cancer patients contained higher levels of chemerin ( Figure 8 ).

[0150] Example 7 Highly Infiltrating FAP + Fibroblasts and SPP1 + Macrophages linked to immunotherapy resistance

[0151] 1. Materials and Methods

[0152] 2. Experimental results

[0153] According to the COAD data in TCGA, the reverse convolution group analysis found that high-infiltration FAP + Fibroblasts and SPP1 + The proportion of lymphocyte infiltration was the lowest in the macrophage group ( Figure 9 A), suggesting that this type of tumor has immune cell rejection characteristics. Current immunotherapy mainly targets lymphocytes, so it is speculated that reduced immune cell infiltration may weaken the efficacy of immunotherapy. To verify this hypothesis, the present invention used the IMvigor210 dataset for PD-L1 treatment and performed a stratified analysis based on the expression of FAP and SPP1. The results showed that the survival of patients in the PD-L1 treatment group with high expression of FAP or SPP1 and the PD-L1 treatment group with high expression of both FAP and SPP1 was shorter ( Figure 9 Importantly, patients with high FAP or SPP1 expression had lower response rates, including complete remission and partial remission ( Figure 9 D) These results suggest that high expression of FAP and SPP1 can reduce the patient's response rate to anti-PD-L1 antibody immunotherapy.

[0154] The protection content of the present invention is not limited to the above embodiments. Without departing from the spirit and scope of the present invention, changes and advantages that can be thought of by those skilled in the art are included in the present invention and are protected by the appended claims.

Claims

1. Use of any one or a combination of FAP and SPP1 as a biomarker of colorectal cancer tumor border cell subpopulation and colorectal cancer biomarker, characterized in that: The application is to prepare a drug for treating colorectal cancer or a reagent for evaluating the prognosis of colorectal cancer by using any one or a combination of FAP and SPP1 as a target; the cell is SPP1 + Macrophages and / or FAP + Fibroblasts.

2. The use according to claim 1, characterized in that Antibodies are prepared by immunizing animals with any one or a combination of FAP and SPP1 as immunogens; and / or immune cells and / or small molecules are prepared with any one or a combination of FAP and SPP1 as targets.

3. The use according to claim 1 or 2, characterized in that The sample for measuring the biomarker is a sample obtained from a patient's tumor tissue or blood.

4. The use according to claim 3, characterized in that The sample for measuring the biomarker is an intestinal tissue or blood sample obtained from the patient.

5. The use according to claim 3, characterized in that The level of any one or more of FAP and SPP1 is determined at the protein level; or the level of any one or more of FAP and SPP1 is determined at the nucleic acid level.

6. The use according to claim 5, characterized in that When the level of any one or more of FAP and SPP1 is determined at the protein level, the level of any one or more of FAP and SPP1 is determined by immunohistochemistry; When the level of any one or more of FAP, SPP1 is determined at the nucleic acid level, the level of any one or more of FAP, SPP1 is determined by quantifying the mRNA encoding any one or more of FAP, SPP1.

7. The use according to claim 1 or 2, characterized in that The lower the level of any one or more of FAP and SPP1, the higher the probability that the patient will have a long survival time.

Images