CN116356018B

CN116356018B - Circulating tumor cell immune check point and application thereof in inhibiting tumor metastasis

Info

Publication number: CN116356018B
Application number: CN202211610754.XA
Authority: CN
Inventors: 石虎兵; 刘小伟; 赵钰洁; 马学磊
Original assignee: West China Hospital of Sichuan University
Current assignee: West China Hospital of Sichuan University
Priority date: 2022-12-14
Filing date: 2022-12-14
Publication date: 2024-05-17
Anticipated expiration: 2042-12-14
Also published as: CN116356018A

Abstract

The invention relates to a Circulating Tumor Cell (CTCs) immune check point and application thereof in inhibiting tumor metastasis, and HLA-E is obtained by screening: a series of specific immune checkpoints between CTCs, such as CD94-NKG2A, and NK cells, mediate immune escape from CTCs. The inhibition of CTCs and NK cells can be blocked by targeting the immune checkpoint, thereby restoring the killing effect of NK cells on CTCs. Blocking the binding of HLA-E on the surface of CTCs and CD94-NKG2A on the surface of NK cells by using an NKG2A binding antibody or inhibiting the expression of HLA-E can enhance the killing effect of NK cells on tumor cells in vitro and effectively inhibit tumor metastasis in vivo. The circulating tumor cell immune check point provides a new treatment target and a treatment scheme for clinical treatment and prevention of tumor metastasis.

Description

Circulating tumor cell immune check point and application thereof in inhibiting tumor metastasis

Technical Field

The invention belongs to the technical field of biological medicines, and particularly relates to a circulating tumor cell immune check point and application thereof in inhibiting tumor metastasis.

Background

Metastasis and spread of tumor cells are the leading cause of cancer-related death. Tumor cells released into the blood circulation from primary focal tumor shedding are considered "seeds" of tumor metastasis, termed circulating tumor cells (Circulating Tumor cells, CTCs). CTCs, which leave the primary tumor foci or metastases and enter the peripheral blood circulation during tumor formation and progression, usually exist in the form of single cells and clusters of cells (two to tens of cells), are considered to be critical in causing tumor metastasis. CTCs are detected in the blood of various tumor species such as metastatic pancreatic cancer, breast cancer, lung cancer, colorectal cancer, prostate cancer, and are closely related to the decrease in tumor progression-free survival (PFS) and total survival (OS). Distant organ metastasis by CTC formation is a major cause of poor prognosis, recurrence of cancer and death in cancer patients. Most CTCs entering the blood circulation are usually deactivated due to shear stress of blood flow, anoikis, recognition and killing of immune cells, and only part of CTCs that have escaped through migration, adhesion and mutual aggregation can form tiny cancer plugs, and finally form metastasis. Thus, research on how CTCs can achieve immune surveillance that evades various immune cells in the blood circulation is an important basis for achieving the blocking of cancer metastasis.

The advent of Immune checkpoint blocking (Immune-checkpoint blockade, ICB) therapy has drastically changed the treatment options for a wide variety of cancer types, now being the first line of choice for current cancer therapies in parallel with radiotherapy, chemotherapy, surgery and molecular targeted drugs. Although ICB therapy has been an unprecedented success, only a few patients have achieved good clinical responses and it is still necessary to find new immune checkpoints and develop new tumor treatment regimens. Currently, a great deal of research has been conducted to explore immune checkpoints between tumor cells and immune cells in the microenvironment of a solid tumor primary or metastatic tumor. CTCs are critical for tumor metastasis and recurrence, however immune checkpoints and immune escape involving CTCs are of little interest due to technical limitations.

Disclosure of Invention

In order to solve the problems in the prior art, the invention provides a circulating tumor cell immune check point and identification and application thereof. The invention combines the micro-fluidic chip technology, the antibody capturing technology and the single-cell transcriptome sequencing technology, analyzes the interaction between the blood circulation and tumor-immune cells in primary and metastatic solid malignant tumors on a single-cell scale by utilizing bioinformatics, and screens out HLA-E: CD94-NKG2A represents immune checkpoint molecule pair closely related to CTC immune evasion, and provides an identification method and application of the circulating tumor immune checkpoint.

The technical scheme adopted by the invention is as follows:

A circulating tumor cell immune checkpoint for immunotherapy, the immune checkpoint molecule being a CTCs-specific immune checkpoint molecule comprising HLA-E: CD94-NKG2A, HLA-E: CD94-NKG2C and HLA-E: one or more of CD94-NKG 2E.

The identification method of the immune check point of the circulating tumor cells comprises the following steps:

(1) Capturing CTCs in blood, and collecting primary and metastatic malignant solid tumor samples to prepare single-cell suspension;

(2) Single cell transcriptome sequencing is carried out on the single cells obtained in the step (1), and immune check point molecules with specificity between CTCs and immune cells are obtained through analysis;

(3) And (3) performing functional verification on the immune checkpoint molecule obtained in the step (2).

The application of the circulating tumor cell immune check point in preparing a medicament for preventing or treating tumor metastasis.

Such tumors include, but are not limited to, pancreatic cancer, melanoma, breast cancer, colorectal cancer, and liver cancer.

Such agents that prevent or treat tumor metastasis include, but are not limited to, small molecule agents, antibody agents, and gene therapy agents.

The small molecule drug is a drug capable of inhibiting the expression level or biological function of the immune checkpoint molecule.

The monoclonal antibody is a binding antibody to the immune checkpoint molecule.

The monoclonal antibody is Mo Nali bead monoclonal antibody.

The gene therapy drug is designed aiming at the immune checkpoint molecule and comprises one or more of DNA drugs, RNA drugs and genetically modified cells.

The gene therapy medicine is one or more of sh-HLA-E, si-HLA-E, sh-NKG2A, si-NKG 2A.

The present inventors have found in long-term studies that, since blood circulation is the primary pathway for CTCs to metastasize to distant organs, studying the interaction between CTCs and immune cells in the blood stream may provide a strategy to block tumor metastasis by activating the host immune elimination system. Further studies have found that analysis of tumor-immune cell interactions in blood circulation and solid malignant lesions on a single cell scale can search for immune checkpoints to obtain CTCs by systematically dissecting immune-related molecular pairs between CTCs and immune cells.

The present inventors analyzed pancreatic cancer primary foci tumors, portal venous blood and liver metastases tumors by single cell transcriptome sequencing. Transcriptome characteristics of primary foci tumor cells, CTCs, and hepatic metastases tumor cells are characterized; the blood circulation and tumor cell-immune cell interactions in solid (primary and metastatic) malignancies were analyzed on a single cell scale using bioinformatics. Through the systematic profiling of immune-related molecule pairs, a specific immune checkpoint between CTCs and NK cells, HLA-E, was found: CD94-NKG2A, which mediates immune surveillance of CTCs against NK cells. In vitro experiments, it was found that blocking the binding of this immune checkpoint directly influences the in vitro killing ability of NK cells against tumor cells. In the in vivo tumor metastasis model of mice, HLA-E is blocked with either a blocking antibody to NKG2A or shHLA-E: CD94-NKG2A immune-checkpoint molecules can effectively inhibit tumor metastasis. Thus the present application identifies a novel immune checkpoint between CTCs and NK cells, HLA-E: blocking this immune checkpoint is effective in inhibiting tumor metastasis by CD94-NKG 2A.

The beneficial effects of the invention are as follows:

1. immune checkpoint HLA-E identified by the invention: CD94-NKG2A is a CTCs-specific immune checkpoint that is highly expressed between CTCs and NK cells of the blood system, but is poorly expressed in primary and metastatic tumor microenvironments.

2. The immune check point HLA-E provided by the invention: CD94-NKG2A exists in CTCs of a plurality of cancer species such as pancreatic cancer, melanoma, liver cancer, breast cancer, colon cancer and the like, and can be widely used for treating distal metastasis of a plurality of tumors.

3. The present invention identifies immune checkpoint HLA-E in patient-derived blood: CD94-NKG2A is present in CTCs and NK cells, which pass through HLA-E: CD94-NKG2A binds.

4. Experiments prove that HLA-E is an immune check point of CTCs: CD94-NKG2A plays a role in evading the immune monitoring of NK cells, and blocking the immune check point can enhance the killing capacity of NK cells to tumor cells.

5. The method for inhibiting tumor metastasis provided by the invention uses specific antibodies and gene therapy drugs. Small molecule drugs and the like interfere the combination of HLA-E on the surface of CTCs and CD94-NKG2A on the surface of NK cells, so that the immunosuppression of the CTCs on the NK cells is relieved, the killing function of the NK cells is recovered, the CTCs are killed efficiently, and the tumor metastasis is effectively inhibited.

6. The invention is achieved by blocking immune checkpoint HLA-E: CD94-NKG2A realizes the prevention and treatment of tumor metastasis, and provides a new target and treatment strategy for inhibiting tumor metastasis by targeting CTCs in clinic.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a diagram of clinical profiles of patient tissue for single cell transcriptome sequencing. Panel A is a CT image of the patient's pancreas, liver and portal vein system. Panel B is a representative image of primary and metastatic focus biopsies during the endoscopic procedure. Panel C is an H & E pathology staining of the primary foci and liver metastases.

FIG. 2 is a graph showing the result of dimension reduction clustering of single cell transcriptome sequencing data, wherein the abscissa and the ordinate represent the first and second principal components after dimension reduction of t-SNE, respectively, and each point represents a different cell. Visualization of cell subsets using a nonlinear t-SNE dimension reduction method.

FIG. 3 is a graph of single cell cluster analysis of primary/metastatic tumors and portal blood of PDACs based on single cell sequencing technology. Panel A is a bubble heat map of typical marker gene expression for each subtype of cell. Panel B shows all sequenced cells based on each cell subset in the t-SNE dimensionality reduction cluster analysis. Panel C is a plot of a t-SNE dimension-reduction cluster analysis of tissue sources of sequenced cells. Panel D is a plot of patient-derived t-SNE dimension-reduction cluster analysis of sequencing cells. FIG. E is a graph of t-SNE dimension-reducing cluster analysis of cell subtypes from primary focal tumors, portal venous blood, and metastatic focal tumors.

FIG. 4 is a graph of immune cell cluster analysis of PDAC portal venous blood, primary and liver metastases based on single cell sequencing technology. Panel A is a t-SNE dimension-reducing cluster analysis of the immune cell subpopulation of CD45 ⁺(PTPRC⁺) in all samples. Panel B is a t-SNE dimension reduction cluster analysis of the tissue sources of each sequenced immune cell. Panel C is a heat map of typical marker gene expression for each cell subpopulation. Panel D shows the expression and distribution of each subtype of cell marker gene in the t-SNE assay.

FIG. 5 is a graph showing the interaction between tumor cells and various subtypes of immune cells in portal blood, primary and hepatic metastases based on CellPhoneD analysis. (A) Interaction between primary tumor cells and immune cells. (B) interaction between CTCs and immune cells. (C) Interaction between metastatic tumor cells and immune cells. The lines in the figure represent the links between cognate receptors or ligands, and the thickness of the lines reflects the number and expression levels of the two intercellular ligand-receptor pairs.

FIG. 6 is a graph of immune checkpoint receptor-ligand analysis between tumor cells and immune cells of each subtype in portal blood, primary and hepatic metastases.

FIG. 7 is a bar graph showing HLA-E expression in CTCs derived from pancreatic cancer liver metastasis, primary and metastatic tumor cells.

FIG. 8 is a bar graph showing HLA-E expression levels in hepatocellular, breast, colon, melanoma, and pancreatic cancer-derived CTCs.

FIG. 9 is a bar graph showing the expression levels of HLA-E in pancreatic cancer mouse-derived CTCs.

FIG. 10 is a bar graph showing the relative expression amounts of KLRC1 (NKG 2A) and KLRD1 (CD 94) in immune cells in portal blood.

FIG. 11 is a graph showing the physical interactions of multiple immunofluorescent staining to detect isolated CTCs with NK in portal blood.

FIG. 12 shows detection of HLA-E and CD94 expression levels in CTCs-NK cell clusters by multiplex immunofluorescent staining.

FIG. 13 is a graph showing the interaction of CTCs with NK in multiple immunofluorescent staining for detection of PDAC liver metastases.

FIG. 14 is a graph of in vitro studies of HLA-E, CD94-NKG2A mediated escape of CTCs from NK immune monitoring. Panel A is a Western blot detection of the expression level of SU86.86 cells HLA-E when HLA-E is knocked out. Panel B is a graph of the expression level of SU86.86 cells HLA-E when the HLA-E is over-expressed by Western blot detection. Panel C is a flow cytometry analysis of NK cells positive for NKG2A and NKG2C in peripheral blood of PDAC patients. Panel D is the quantification of panel C. FIG. E is an LDH release experimental graph, and the killing capacity of NK to SU86.86 cells on tumor cells under the condition of different effective target ratios (E: T) is detected. FIG. F shows the expression of NK cytotoxic effector GZMB and immune checkpoint related kinase SHP1 after Western blot detection of NK cells co-incubated with shHLA-E SU 86.8.

FIG. 15 is a graph showing the results of in vivo studies of the effect of blocking NKG2A with an NKG 2A-binding antibody on lung metastasis in mice. And A is an experimental design scheme. Panel B shows that mice were examined for lung metastasis growth by bioluminescence in vivo imaging 15 days after intravenous inoculation of KPC-Luc cells. Panel C is a statistical quantification of lung metastases from panel B. Panel D is a photograph of a lung metastasis nodule. Panel E is a lung H & E staining pattern.

FIG. 16 shows the expression of H2-T23 in KPC cells with shH-T23 knockdown by Western blot.

FIG. 17 is a graph showing the results of in vivo studies in which the expression of knockdown HLA-E reduces lung metastasis in mice. Panel A is a bioluminescence image of lung metastasis of Balb/c nude mice after injection of KPC-Luc cells knocked out of H2-T23. Panel B is a quantitative plot of A. Panel C is a representative visual image of the lungs of representative mice of each group. Panel D is a lung tumor nodule count. Panel E is an H & E staining of the lung.

FIG. 18 is a C57BL/6 mice in vivo blocking immune checkpoint HLA-E: results of CD94-NKG2A inhibition of lung metastasis in mice. Figure a is a design of a mouse lung metastasis protocol. C57BL/6 mice were either intravenously inoculated with 5X 10 ⁴ KPC-Luc/H2-T23 knocked-out KPC-Luc cells or subjected to blocking treatment by injection of anti-NKG2A antibody (10 mg/kg) prior to inoculation. Panel B is a lung metastasis image of mice detected by in vivo imaging of the mice 15 days after intravenous injection of KPC-Luc cells. Panel C is a quantitative statistical plot of Panel B. Panel D is a photograph of mouse lung tissue. And E is a lung metastasis node number statistical graph. Panel F is an H & E stained photograph of lung tissue.

FIG. 19 is a block immune checkpoint HLA-E: results of CD94-NKG2A studies showing that mice have prolonged survival.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It will be apparent that the described embodiments are only some, but not all, embodiments of the invention. All other embodiments, based on the examples herein, which are within the scope of the invention as defined by the claims, will be within the scope of the invention as defined by the claims.

The invention takes pancreatic cancer liver metastasis as a research object, and uses single cell transcriptome sequencing to analyze pancreatic primary focus tumor, portal vein blood and liver metastasis tumor. Transcriptome characteristics of pancreatic cancer primary foci tumor cells, CTCs and hepatic metastases tumor cells were characterized, and the interaction between blood circulation and tumor-immune cells in primary and metastatic solid malignancies was analyzed on a single cell scale using bioinformatics. Through the systematic profiling of immune-related molecule pairs, a specific immune checkpoint HLA-E, CD94-NKG2A, was found between CTCs and NK cells that mediated immune surveillance of CTC immune evasion NKs.

In vitro experiments, it was found that overexpression or knockout of HLA-E expression of tumor cells directly affects NK cell killing ability in vitro on tumor cells. In a mouse in vivo tumor metastasis model, HLA-E is blocked by using a blocking antibody of NKG2A or shHLA-E, CD94-NKG2A immune check point molecules can effectively inhibit tumor metastasis, which shows that HLA-E CD94-NKG2A is a novel immune check point between CTC and NK cells.

The following describes the above technical scheme in detail with reference to specific embodiments.

The experimental methods used in the following examples are conventional methods unless otherwise specified. Materials, reagents and the like used in the examples described below are commercially available unless otherwise specified.

Example 1 Primary focal tumor, metastatic tumor and CTCs single cell transcriptome sequencing and banking

1. Sample collection

To find immune checkpoints specific for CTCs, we take pancreatic cancer liver metastasis as an example, primary foci, CTCs and metastatic foci of the same patient were collected for single cell transcriptome sequencing. First, we screened 6 untreated PDAC liver metastasis patients by computer tomography (Computed Tomography, CT) tumor markers (fig. 1A). Then, the patient's hepatic portal venous blood (hepatic portal vein, HPV), the living tissues of primary and liver metastatic lesions (fig. 1B) were collected by laparoscopic surgery, and histopathological verification was performed to confirm that the tissues taken were pancreatic primary foci tumor and metastatic foci tumor (fig. 1C). Primary foci and metastatic foci tumor tissues obtained in the operation are temporarily stored in tissue preservation solution and placed on ice, and then immediately sent to a laboratory for subsequent dissociation of single-cell samples and single-cell transcriptome sequencing; portal blood is stored in anticoagulation blood collection tubes and sent to the laboratory for subsequent isolation of CTCs and single cell transcriptome sequencing.

2. Preparation of single cell samples

The primary and metastatic tumor tissues were minced on ice with scissors. Subsequently, the tissue pieces were transferred to a 50ml centrifuge tube, and 20ml of a digestive juice was added thereto for digestion at 37℃for 15min; wherein the digestion solution contains 0.25% trypsin, 0.4mg/ml type I collagenase and 0.4mg/ml type IV collagenase. After complete digestion of the tumor tissue, the reaction was stopped with an equal volume of pre-chilled DMEM medium containing 10% fetal bovine serum, and the bulk debris was removed with a 70 μm cell screen and the single cell suspension was collected. Next, the single cell suspension was centrifuged at 500g for 5min at 4 ℃. Adding the centrifuged cell precipitate into erythrocyte lysate for cracking for 5min; the cell pellet was then collected by centrifugation at 500g at 4℃twice with HBSS (0.1% BSA). Then, the cells were resuspended with HBSS containing 0.1% bsa and PI was added and the live cells with good flow cytometry sorting status were used for single cell transcriptome sequencing.

3. Preparation of scRNA-seq library

After completion of the preparation of the single cell samples described above, single cell transcriptome libraries were prepared using a 10× Genomics Chromium 3'Gene Expression Kit V3 kit and sequenced. Specifically, target cells and corresponding 10×genomics reagents are added to a chromanum chip to generate latex droplets (Gel Bead in Emulsion, GEM) containing single cells and single gel beads, and subsequent reverse transcription, double-stranded DNA generation, library construction, and sequencing are performed. In the laboratory, we set the target capture cells for each sample to be 6,000-8,000. The final constructed library was sequenced on an Illumina HiSeq 4000 platform with a target sequencing depth of 100,000 reads per cell.

Example 2 identification of CTCs-specific immune checkpoint molecules

(1) Single cell transcriptome data preprocessing and quality control

The raw sequencing data image obtained by the sequencing is firstly converted into Fastq format sequence information with 150 base pairs at double ends by using the bcl2fastq software of Illumina company. The resulting sequences were then aligned with human reference genome version GRCh38 using CELL RANGER (v.3.0.0, 10 x Genomics) software to obtain a gene expression matrix. UMI tags that can specifically compare sequences to the exon regions of transcriptome genes will be used for subsequent statistical counting. The Seurat R software package then performs subsequent quality control, we filtered the cells and genes by three criteria: (1) Cells with gene expression numbers exceeding 7500 and less than 200 are excluded, because low quality cells typically have a lower number of gene expressions, while dual or multicellular gene expression numbers are typically higher; (2) Cells with more than 25% of mitochondrial gene numbers were removed, as dead cells generally showed high mitochondrial contamination; (3) excluding genes expressed in less than three cells. In addition, cells with a variety of different types of cell marker genes were identified as multicellular. After quality control, a total of 74,206 cells were measured for 18 samples, including 27,296 cells from primary focal tumors, 6,922 cells from liver metastases, and 9,988 cells from portal blood.

(2) Single cell transcriptome data dimension reduction and cluster analysis

After quality control of cells and genes, single cell transcriptome data was normalized using Seurat (v.4.0.1) R software package, respectively, to eliminate differences between the cell data and to perform subsequent reduced and unsupervised cluster analysis. Firstly, we use global scaling normalization method 'LogNormalize' to globally normalize the gene expression matrix of single cell collection, and take the log value after dividing the expression value of the gene by the total expression value of the corresponding single cell and multiplying the product by the parameter factor (using default standard parameter 10000). Next, the first 2000 high variable genes were selected using the FindVariableFeatures function and the normalized data was scaled to z-score using the SCALEDATA function. Then, the high-variable genes are subjected to principal component analysis (PRINCIPAL COMPONENTS ANALYSIS, PCA) by utilizing RunPCA functions, and the high-dimensional data are subjected to dimension reduction processing. After completion of dimension reduction, a KNN map was constructed using FindNeighbors functions to define the weights between two cells. The Louvain algorithm is then applied to group similar cells by FindClusters functions, with the resolution (resolution) parameter set to 1. The single cell transcriptome data was then subjected to dimensionality reduction cluster analysis using a non-linear t-SNE dimensionality reduction method (fig. 2).

(3) Copy number variation analysis and identification of cell types

Cell types were identified using the method of sciBet (v.1.0) R software package, classical marker gene and copy number variation in combination. First, cell types were annotated according to classical cell markers, and total of 29,930 epithelial cells (EPITHELIAL: EPCAM ⁺、KRT8⁺、KRT18⁺、KRT19⁺), 1,675 fibroblasts (fibre bribolast: FAP ⁺、COL1A1⁺、DCN⁺), 875 Endothelial cells (endothesil: VWF ⁺、CDH5⁺、ENG⁺、PECAM1⁺), 523 CTCs (PTPRC ^-、PPBP⁺、PF4⁺、CD9⁺、KRT8⁺、TIMP1⁺), 19,072 myeloid cells (Myeloid: AIF1 ⁺、CD14⁺、LYZ⁺、FPR1⁺), 1,448B cells (B cells: CD79A ⁺、CD79B⁺、MS4A1⁺), 2,485 NK cells (NK cells: KLRF1 ⁺、KLRD1⁺、GNLY⁺), 18,198T cells (T cells: CD3D ⁺、CD3E⁺、CD3G⁺) were noted (FIG. 3A). Then, we performed t-SNE dimension-reduction clustering and visual analysis of all single cells based on cell type, tissue source and patient source (FIGS. 3B-D). And differences in cell types in primary focal tumor tissue, portal blood and metastatic focal tumor tissue were analyzed using t-SNE dimension reduction (FIG. 3E).

Whereas we performed separate t-SNE dimension-reducing cluster analyses on immune cells from portal venous blood, primary and hepatic metastases tumor tissue PTPRC ⁺(CD45⁺ (FIGS. 4A, B). Annotation of all subtypes of immune cells according to classical immune cell markers including NK cells (NKs; KLRD1 ⁺、KLRF1⁺), NK-T cells (CD 3D ⁺、CD3E⁺、KLRD1⁺、KLRF1⁺) CD8 depleting T cells (CD 8 Ex; CD8A ⁺、PDCD1⁺、LAG3⁺), CD8 effector T cells (CD 8 EFF; CD8A ⁺、IFNG⁺、GZMA⁺), memory T cells (Memory T; CD3D ⁺、CD44⁺、IL7R⁺、LTB⁺), naive T cellsT is a T; CD3D ⁺、CCR7⁺、TCF7⁺、SELL⁺、LEF1⁺), treg cells (CD 4 ⁺、FOXP3⁺、IL2RA⁺), B cells (CD 79a ⁺、CD79B⁺), neutrophils (Neutrophile; CD14 ^-、FCGR3B⁺、FPR1⁺), monocytes (Monocyte; CD14 ⁺、S100A12⁺、FCGR3A⁺), M1 macrophages (M1 macrophage; FCGR3a ⁺、CD68⁺、ITGAX⁺、ITGAM⁺), M2 macrophages (M2 macrophage; CD163 ⁺、MRC1⁺、MSR1⁺), classical DC cells (cDC; CD1C ⁺、FCER1A⁺、CLEC10A⁺), plasmacytoid DC (pDC; LILRA4 ⁺、IL3RA⁺) and mast cells (mass cells; MS4A2 ⁺、TPSAB1⁺、KIT⁺) (fig. 4c, d).

(4) Analysis of cell-cell interactions

To study the interactions between tumor cells and immune cells during tumor metastasis, we used CellPhoneDB to analyze the interactions of tumor cells with ligand-receptor pairs on the surface of immune cells of each subtype based on gene expression levels and differences. The CellPhoneDB database consisted of 1396 ligand molecule pairs from UniProt, ensembl, PDB, IMEx and iuphas databases. Molecular pairs of potential interactions between tumor cells and immune cells are obtained by extracting expression of related genes from single cell transcriptome data and comparing with a reference database. Wherein the molecular pairs of interactions between different types of cells are determined by: 1) Integrating the expression analysis of the ligand and receptor between each cell type to calculate the average expression level of the ligand; 2) And calculating the p value of the expression change by using EMPIRICAL SHUFFLING algorithm, and screening out the ligand molecule pair with different change. After the ligand molecule pairs of different types of cell interactions are obtained by the method, the ligand molecule pairs are classified according to functions of the ligand molecule pairs by utilizing CellChat database. We found that ligand-receptor pairs that interact between primary foci of tumor cells, CTCs, metastatic focus of tumor cells and immune cells were significantly different, indicating that immune cell monitoring of tumor cells was dynamic during tumor metastasis (fig. 5A-C). In primary and metastatic tumor microenvironments, there is a complex and strong interaction between tumor cells and various immune cells such as CD8 ⁺ T cells, macrophages, NKs, etc. While the interaction between CTCs and immune cells in the blood circulation is relatively simple, CTCs interact primarily with NK cells in the blood, suggesting that NK cells may be involved in immune monitoring of CTCs in the blood circulation.

(5) Identification of immune checkpoints specific for CTCs

We first screened all immune checkpoint molecule pairs in CellPhoneDB, then calculated the average expression levels of the ligand molecule pairs in tumor cells (including CTCs) and immune cells of each subtype, calculated p-values of the expression changes using EMPIRICAL SHUFFLING algorithm, and calculated immune checkpoint molecule pairs present between primary, CTC, metastatic and immune cells of each subtype. We found that, in marked contrast to primary and metastatic tumor cells, there are some specific pairs of immune checkpoint molecules between CTCs and immune cells in the blood circulation (fig. 6). Several pairs of highly expressed, specific immune checkpoint molecules were found between CTCs and NK cells, including HLA-E where CTCs interact with NK: KLRC1 (NKG 2A), HLA-E: KLRC2 (NKG 2C) and HLA-E: KLRK1 (NKG 2D), and CD94-NKG2A where NK interacts with CTCs: HLA-E, CD94-NKG2C: HLA-E and CD94-NKG2E: HLA-E.

Example 3 verification of the Presence of HLA-E: CD94-NKG2A immune-checkpoint molecules between CTCs and NK cells

(1) Detection of HLA-E expression levels in CTCs

First, we calculated the expression levels of HLA-E in primary tumor cells, CTCs in blood and metastatic tumor cells after Log2 (tpm+1) normalization of gene expression values using the above CTCs transcriptome sequencing data and with primary tumor cells and metastatic tumor cells as controls. The results are shown in FIG. 7, which shows that HLA-E is highly expressed in CTCs compared to primary and metastatic tumor cells, suggesting that it may play a potential role in the metastasis of CTCs.

At the same time, we also examined the expression level of HLA-E in other tumor-derived CTCs. We calculated the expression levels of HLA-E in CTCs of different tumor origin using transcriptome data of hepatocellular carcinoma (HCC; CNP0000095, GSE 117623), PDAC (GSE 144561), breast cancer (BRCA; GSE67939, GSE 86978), colon adenocarcinoma (COAD; GSE 74369) and melanoma (SKCM; GSE 38495) origin CTCs in the database, and normalizing the gene expression values with Log2 (tpm+1). The results are shown in FIG. 8, where HLA-E is highly expressed in hepatocellular carcinoma, pancreatic carcinoma, breast carcinoma, colon adenocarcinoma, and melanoma-derived CTCs, indicating broad spectrum applicability of the immune checkpoint molecule.

In addition, we examined the expression level of HLA-E in other mouse tumor-derived CTCs, and in mice the homologous gene of HLA-E was H2-T23, so that all the following references to H2-T23 refer to HLA-E. We calculated the expression level of H2-T23 (HLA-E) in mouse tumor-derived CTCs using transcriptome data from mouse pancreatic cancer (GSE 51372) in the database and normalizing the gene expression values with Log2 (tpm+1). As shown in FIG. 9, H2-T23 (HLA-E) was also highly expressed in mouse tumor-derived CTCs.

(2) Detection of expression levels of CD94 and NKG2A in NK cells

Immune checkpoint molecule pairs HLA-E were detected using single cell sequencing data: ligand molecules NKG2A (KLRC 1) and CD94 (KLRD 1) of CD94-NKG2A are expressed in immune cells of each subtype, such as monocytes, NK cells, B cells, DC cells, T cells, neutrophils, etc. As shown in fig. 10, NKG2A (KLRC 1) and CD94 (KLRD 1) were specifically highly expressed in NK cells, indicating that the ligand molecules of the immune checkpoint were mainly expressed in NK; HLA-E in CTCs binds to CD94-NKG2A on the surface of NK cells. The immune checkpoint molecule is present between CTCs and NK cells.

(3) CTCs were demonstrated to bind to NK cells to form CTCs-NK cell clusters in portal blood of patients with pancreatic cancer liver metastases.

To verify the interaction of CTCs with NK cells, CTCs in portal blood of patients with pancreatic cancer liver metastasis were captured using microfluidic chips. After elution of CTCs, epCAM/CD94 immunofluorescent staining was used to detect the presence of CTC-NK cell clusters in the eluate. As a result, as shown in FIG. 11, CTC-NK cell clusters were observed in the eluents of CTCs from multiple patients, indicating direct interactions between CTCs and NK cells.

(4) Immune checkpoint molecule HLA-E was demonstrated by HLA-E/CD94 multiplex immunofluorescent staining: CD94-NKG2A exists between CTCs and NK cells.

To verify immune checkpoint HLA-E: CD94-NKG2A exists between CTCs and NK cells. We captured CTCs in portal blood of PDAC liver transfer patients using microfluidic chips. After elution and collection of CTCs, HLA-E/CD94 immunofluorescent staining was used to detect whether the CTCs-NK cell cluster expressed HLA-E and CD94. The results are shown in FIG. 12, where HLA-E and CD94 protein expression was detected in CTCs and NK cells, respectively, indicating that this immune checkpoint molecule pair exists between CTCs and NK cells.

(5) The interacting CTCs-NK cell clusters were detected in patients with pancreatic cancer liver metastases.

We collected and examined pancreatic cancer hepatic metastasis patient primary foci and hepatic metastasis tumor tissues, and stained patient tissues by EpCAM/CD94 immunofluorescence for the presence of CTCs-NK cell clusters. As shown in fig. 13, we detected co-localization of CTCs with NK in a healthy liver tissue region near the metastasis; this result further demonstrates the interaction between CTCs and NK cells.

Example 4 blocking immune checkpoint HLA-E CD94-NKG2A enhances the killing ability of NK cells against tumor cells

In this example, we assessed the effect of the intervention immune checkpoint HLA-E: CD94-NKG2A on the ability of NK cells to kill tumor cells.

Blocking this immune checkpoint binding in two ways:

1) HLA-E protein is pre-expressed or knocked out in human PDAC cells SU86.86 by lentivirus over-expression and shRNA interference technology, and the binding of the HLA-E protein and NK cell surface NKG2A is interfered. For the expression of HLA-E of human tumor cells by shRNA interference technology, we designed shRNA sequences (table 1) targeting human HLA-E (shH-T23), cloned to pLKO.1-puro vector, extracted plasmids, then introduced into tumor cells by lentiviral infection technology to knock out the expression of HLA-E, and detected by Western blot, and as a result, the designed shRNA sequences effectively knock down the expression of HLA-E of human tumor cells as shown in FIG. 14A. For over-expression of HLA-E protein, we synthesized HLA-E gene sequence, cloned to Plvx-puroz plasmid vector, extracted plasmid, then introduced into tumor cells to express HLA-E protein by using lentiviral infection technique, and detected by Western blot, and the result shows that the expression level of HLA-E in human tumor cells is significantly up-regulated after over-expression of HLA-E, as shown in FIG. 14B.

TABLE 1 sequences for construction of shRNA expression vectors

2) NKG2A of which the final concentration is 100 mug/ml Mo Nali bead monoclonal antibody (monalizumab) for blocking NK cells is added into a co-culture system of tumor cells and NK cells to block the combination of the NK cells and tumor cells HLA-E.

First, NK cells were isolated from peripheral blood of PDAC patients, and after flow cytometry, NK cells derived from PDAC patients were found to highly express NK inhibitory protein NKG2A and significantly higher than activating protein NKG2C (shown in FIGS. 14C and D, the ratio of NKG2A and NKG2C in peripheral blood NK cells of PDAC patients was examined and quantified by flow cytometry.

NK cells were then co-incubated with SU86.86 tumor cells at different potency target ratios (5:1 and 10:1), after 24 hours of co-incubation, cell supernatants were collected and assayed for NK cell killing by LDH kit. The results are shown in FIG. 14E, where the LDH release assay detects NK's ability to kill tumor cells under different effective target ratio (E: T) conditions for SU86.86 cells. Data are shown in mean±se, n=3, p <0.05, p <0.01, t-test.

SU86.86 over-expressed HLA-E protected from NK cell killing at both different potency ratios. In contrast, the NK cell killing capacity of tumor cells is remarkably enhanced by utilizing shHLA-E to knock out the expression of SU86.86 cell HLA-E. Similarly, blocking of NKG2A with Mo Nali bead mab also enhanced NK cell killing of SU86.86 and HLA-E overexpressing SU 86.86.

Example 5

In this example, WB verifies blocking CD94-NKG2A: HLA-E immune checkpoints increase NK activity.

Empty vector and shHLA-E transfected SU86.86 tumor cells were seeded at a density of 5X 10 ⁴ cells/well in 24 well plates and cultured overnight. The next day, NK cells were added in a ratio of 10:1 for co-incubation for 24 hours. NK cells were then collected and the activity of SHP-1 downstream of the NKG2A immune checkpoint was examined by western blot (Western blot) and the GZMB protein expression level that marks the NK cell activity. As shown in FIG. 14, after NK cells were co-incubated with shHLA-E SU86.8, western blot detected the cytotoxic effector of NK, GZMB, and the immune checkpoint related kinase, SHP1.

When HLA-E knockdown tumor cells were incubated with NK, the activity of SHP-1 downstream of the NKG2A immune checkpoint in NK cells was inhibited, while the expression of GZMB, which marks the activity of NK cells, was up-regulated (FIG. 14F). The above results indicate that tumor cells can pass through immune checkpoint HLA-E: CD94-NKG2A evades immune surveillance of NK cells.

Example 6

In this example, blocking of immune checkpoint HLA-E with anti-NKG2A antibodies was studied: effect of CD94-NKG2A on tumor metastasis in mice.

Next, blocking immune checkpoint HLA-E was studied in mice: whether CD94-NKG2A inhibited metastasis of tumor. Since Balb/c nude mice are immunodeficiency mice containing NK cells and without T cells, the nude mice are ideal models for researching NK cells and tumor cells; whereas the mouse PDAC KPC cells endogenously express H2-T23 proteins homologous to human HLA-E proteins (hereinafter, mouse H2-T23 corresponds to human HLA-E genes). Thus, in Balb/c nude mice as a model, tumor cell metastasis was reproduced by tail vein injection of luciferase-labeled KPC (KPC-Luc) cells.

To block tumor cells, which indicate binding of HLA-E to NK cell surface CD94-NKG2A, this immune checkpoint was blocked by two means: (1) blocking NK cell surface NKG2A with a blocking antibody; (2) mice were vaccinated with KPC cells that had been pre-knocked out H2-T23.

First, blocking NK cells with anti-NKG2A blocking antibodies suggests NKG2A. Specifically, 4 doses of anti-NKG2A antibody (10 mg/kg) are respectively injected into continuous tail veins before (day-1), during (day 0) and after (day 1,3 and 5) KPC-Luc cell inoculation for blocking treatment, so as to study the influence of blocking time on the growth of lung metastasis tumor; mice were intraperitoneally injected with potassium D-fluorescein (150 mg/kg) 15 days after inoculation, and mice were subjected to in vivo imaging to detect the intensity of pulmonary fluorescence signal for monitoring growth of lung metastasis (FIG. 15A).

The results show that anti-NKG2A has time dependence on inhibiting tumor metastasis, and the earlier the NKG2A is blocked, the better the effect of inhibiting tumor metastasis. Injection of anti-NKG2A antibody almost completely inhibited tumor metastasis before tumor inoculation (day-1); whereas tumor metastasis is only partially inhibited at day0 and day 1. NKG2A blockade was performed 3 days and 5 days after inoculation, with little inhibition of tumor metastasis (fig. 15b, c).

After the experiment is finished, collecting lung tissues of the mice, photographing, and counting the number of tumor nodules on the lung surface.

The results showed that, consistent with the trend of fluorescence signal quantification, almost no tumor nodules were observed in lung tissue of the anti-NKG2A antibody group injected prior to tumor inoculation (day-1); only a small number of lung metastasis nodules were observed in day0 and day1 groups; whereas day3 and day5 groups of lung metastasis nodules were comparable to the untreated control group (fig. 15d, e). Further H & E pathology analysis was performed on lung tissue, with results consistent with fluorescent signal and lung metastasis nodule count.

The above results indicate that immune checkpoint HLA-E is blocked with anti-NKG 2A: the anti-tumor effect of CD94-NKG2A is only effective against tumor cells in the blood circulation, but has no significant inhibitory effect on tumor cells that have been colonized in the metastatic organ.

Example 7

In this example, it was verified whether inhibition of tumor cell HLA-E expression, blocking of its binding to NK cell surface NKG2A, could inhibit tumor.

Specifically, we designed shRNA sequence (Table 1) targeting mouse HLA-E (shH 2-T23), cloned to pLKO.1-puro vector, extracted plasmid, then introduced into KPC cell by slow virus infection technique to knock out H2-T23 expression, and detected H2-T23 expression by Western blot.

The results showed that the designed shRNA sequence significantly reduced the expression of KPC cell H2-T23, with shH-T23-2 knockdown being most efficient (FIG. 16). Therefore, shH-T23 KPC cells were constructed using this sequence in the experiments, and subsequent animal experiments were performed.

KPC cells stably expressing shH-T23 were then inoculated into nude mice by tail vein injection. Similar to the anti-NKG2A antibody blocking results, decreasing the expression of KPC cell surface H2-T23 (HLA-E) significantly alleviated the formation of tumor cell lung metastases (FIGS. 17A, B).

After the experiment is finished, collecting lung tissues of the mice, photographing, and counting the number of tumor nodules on the lung surface. The results showed that lowering H2-T23 significantly reduced the number of pulmonary tumor nodules with few tumors observed in the shH-T23 group of lung tissues (FIGS. 17C, D). At the same time, the lung tissue was subjected to H & E staining for pathological analysis, and the results were consistent with morphological observation, and formation of tumor metastasis was hardly detected even at the pathological level (FIG. 17E)

Example 8

In this example, it was confirmed that blocking immune checkpoint HLA-E, CD94-NKG2A, also inhibited mouse metastasis in immunized complete mice C57/BL6 mice.

To further verify the role of this immune checkpoint molecule in fully immunized mice, mice lung transfer model was constructed using C57/BL6 immunized fully mice tail vein inoculated with 5 x 10 ⁴ KPC-Luc cells; and blocks this immune checkpoint by two means: (1) Four doses of anti-NKG2A antibody (10 mg/kg) were continuously injected intravenously to block NK cell surface NKG2A starting the day before inoculation (day-1); (2) Inoculation modeling was performed with KPC cells pre-knocked out H2-T23 (FIG. 18A).

Mice were examined for lung metastatic growth using in vivo imaging of the mice 15 days after inoculation.

The results showed that both anti-NKG2A pre-blocking and shH2-T23 pre-knockout significantly inhibited lung metastasis formation (fig. 18B). Quantification of pulmonary fluorescence signals revealed that anti-NKG2A and shH2-T23 reduced pulmonary metastasis fluorescence signals by 63-fold and 115-fold, respectively (FIG. 18C). Lung nodule counts and pathology analysis further showed that anti-NKG2A and shH2-T23 significantly reduced the formation of lung metastases in mice compared to untreated control (fig. 18D-F).

Example 9

In this example, blocking immune checkpoint HLA-E, CD94-NKG2A, was examined to increase survival in mice.

To explore the effect of blocking immune checkpoint HLA-E: CD94-NKG2A on survival of mice, we used Balb/c nude tail veins to inoculate KPC cells (5X 10 ⁴/mice) in another independent cohort, and constructed a mouse lung metastasis model; and blocks this immune checkpoint by two means: (1) Four doses of anti-NKG2A antibody (10 mg/kg) were continuously injected intravenously to block NK cell surface NKG2A starting the day before inoculation (day-1); (2) Inoculation modeling was performed with KPC cells pre-knocked out of H2-T23. Then, the survival of the mice was observed, and a Kaplan-Meier survival curve of the mice was drawn. As shown in FIG. 19, blocking NKG2A with anti-NKG2A antibody and decreasing the expression of tumor cell HLA-E (H2-T23) significantly prolonged the survival of lung metastatic mice.

The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. Use of a blocker of immune checkpoints of circulating tumor cells for immunotherapy in the preparation of a medicament for preventing tumor metastasis, characterized in that the immune checkpoint molecules are CTCs-specific immune checkpoint molecules HLA-E: CD94-NKG2A;

The tumor is pancreatic cancer;

The blocking agent is Mo Nali bead monoclonal antibody for blocking CD94-NKG2A or shHLA-E for reducing HLA-E expression;

the shHLA-E is selected from any pair of sequences:

（1）shHLA-E-1F：CCGGCACCTCTGTGTCTACCATGACCTCGAGGTCATGGTAGACACAGAGGTGTTTTTG；

shHLA-E-1R：AATTCAAAAACACCTCTGTGTCTACCATGACCTCGAGGTCATGGTAGACACAGAGGTG；

（2）shHLA-E-2F：CCGGGTGTTCCTTCCCTGTTCTCTTCTCGAGAAGAGAACAGGGAAGGAACACTTTTTG；

shHLA-E-2R：AATTCAAAAAGTGTTCCTTCCCTGTTCTCTTCTCGAGAAGAGAACAGGGAAGGAACAC；

（3）shHLA-E-3F：CCGGCCTTGAAGTATTTCCACACTTCTCGAGAAGTGTGGAAATACTTCAAGGTTTTTG;shHLA-E-3R：AATTCAAAAACCTTGAAGTATTTCCACACTTCTCGAGAAGTGTGGAAATACTTCAAGG;

（4）shH2-T23-1F：CCGGTCCTGGACCGCGAATGACATACTCGAGTATGTCATTCGCGGTCCAGGATTTTTG；

shH2-T23-1R：AATTCAAAAATCCTGGACCGCGAATGACATACTCGAGTATGTCATTCGCGGTCCAGGA；

（5）shH2-T23-2F：CCGGACATAGCCTCACAGATCTCTACTCGAGTAGAGATCTGTGAGGCTATGTTTTTTG；

shH2-T23-2R：AATTCAAAAAACATAGCCTCACAGATCTCTACTCGAGTAGAGATCTGTGAGGCTATGT；

（6）shH2-T23-3F：CCGGAGATCTCTAAGCACAAGTCAGCTCGAGCTGACTTGTGCTTAGAGATCTTTTTTG；

shH2-T23-3R：AATTCAAAAAAGATCTCTAAGCACAAGTCAGCTCGAGCTGACTTGTGCTTAGAGATCT。