Application of effector memory T cell as marker for predicting TNBC chemotherapy effect

Abstract

The invention relates to the field of biomedical detection. The invention discloses application of a ZNF683 + effect memory CD8T cell as a marker for predicting the curative effect of TNBC chemotherapy and in preparation of a kit for predicting the curative effect of TNBC chemotherapy. Compared with the prior art, the invention discloses that the ZNF683 + CD8T cell is used as the marker for the first time, the kit is prepared, and the chemotherapy curative effect of a patient is predicted by detecting the level of the ZNF683 + CD8T cell in the blood of the TNBC patient before and after treatment. The invention provides a high-reliability, simple and fast way for predicting and evaluating the chemotherapy curative effect of TNBC, especially the curative effect of neoadjuvant chemotherapy.

Description

Technical Field

[0001] The present invention relates to the field of biomedical detection, and in particular to the use of effector memory T cells as markers for predicting the efficacy of TNBC chemotherapy. Background Art

[0002] Triple-negative breast cancer (TNBC) is highly malignant and insensitive to endocrine therapy. Once chemotherapy resistance occurs, there will be a lack of effective treatment. Although the research on the tumor immune microenvironment (TME) of TNBC has gradually increased in recent years, there are still few reports on the effect of neoadjuvant chemotherapy (NAC) on TNBC. Tissue-resident memory T cells (TRM) can respond quickly to tumor antigens and play an important role in the body's anti-tumor immunity. Tumor-draining lymph nodes (TdLN) serve as T cell reservoirs and can deliver effector T cells to cancer foci. However, it is still unknown whether TRM can also be delivered to cancer foci via TdLN, and it is also unknown whether chemotherapy can promote this delivery. Recently, in a study on the effect of neoadjuvant chemotherapy on immune cells in TME and TdLN of TNBC patients, the inventors found that TdLN can also deliver TRM to cancer foci and promote the efficacy of chemotherapy. Summary of the invention

[0003] In a study on the effect of neoadjuvant chemotherapy on immune cells in TME and TdLN of TNBC patients, the inventors found that the TCRs of all CD8 T cells in breast lesion samples of the SD group and the CR group were projected onto the UMAP images of CD8 T cells in blood samples of the corresponding groups. The results showed that compared with the SD group, the TCRs of CD8 T cells in breast lesion samples of the CR group were more significantly enriched in ZNF683 in the blood CD8 T cell samples. + In the same way, the TCR of CD8 T cells amplified in breast lesion samples of the SD group and the CR group was projected onto the UMAP image of CD8 T cells in the blood samples of the corresponding groups. The results also showed that compared with the SD group, the TCR of CD8 T cells in breast lesion samples of the CR group was more significantly enriched in ZNF683 in the blood CD8 T cell samples. + CD8 T cells. And ZNF683 in tumors during neoadjuvant chemotherapy in TNBC + The number of effector memory CD8 T cells is positively correlated with the efficacy of chemotherapy. This suggests that the effector memory T cell ZNF683 in the blood of TNBC patients before and after treatment + The level of CD8 T cells can serve as a biomarker to predict the patient's response to chemotherapy treatment.

[0004] Based on the above findings, the present invention discloses the following technical solutions:

[0005] 1. The use of ZNF683+ effector memory CD8 T cells as a marker for predicting the efficacy of chemotherapy in triple-negative breast cancer.

[0006] 2. Use of ZNF683+ effector memory CD8 T cells in the preparation of a kit for predicting the efficacy of chemotherapy for triple-negative breast cancer.

[0007] The kit includes substances for qualitatively and / or quantitatively detecting ZNF683+ effector memory CD8 T cells.

[0008] The substances for qualitative and / or quantitative detection include substances for identifying the cell marker gene ZNF683.

[0009] The use includes identifying ZNF683 by single cell sequencing + Effector memory CD8 T cell approach.

[0010] ZNF683 + Effector memory CD8 T cells were obtained from the blood of patients with triple-negative breast cancer.

[0011] The present invention discloses for the first time that ZNF683 + CD8 T cells were used as markers, and a kit was prepared to detect ZNF683 in the blood of TNBC patients before and after treatment. + The level of CD8 T cells predicts the efficacy of chemotherapy in patients. + If the level of CD8 T cells is significantly higher than that before treatment, it can be estimated that the chemotherapy effect is better. The present invention provides a highly reliable, simple and fast way to predict and evaluate the chemotherapy of TNBC, especially the efficacy of neoadjuvant chemotherapy. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a composition diagram of CD8 T cells in breast lesion samples of CR and SD groups in the experiment of the present invention; wherein Part A is a sample composition diagram of CD8 T cells in breast lesions of the CR cohort; Part B is a subtype classification diagram of CD8 T cells in breast lesions of the CR cohort; Part C is a landmark subtype gene diagram of CD8 T cells in breast lesions of the CR cohort; Part D is a sample composition diagram of CD8 T cells in breast lesions of the SD cohort; Part E is a subtype classification diagram of CD8 T cells in breast lesions of the SD cohort; Part F is a landmark subtype gene diagram of CD8 T cells in breast lesions of the SD cohort.

[0013] Figure 2Figure 3 is a diagram showing the classification of CD8 T cells in the draining lymph nodes of the CR group and the changes of each subtype in CD8 T cells before and after treatment; Part A is a diagram showing the composition of each sample in the draining lymph node CD8 T cells of the CR cohort; Part B is a diagram showing the subtypes of CD8 T cells in the draining lymph node samples of the CR cohort; Part C is the expression of each marker gene of CD8 T cells in the draining lymph node samples of the CR cohort; Part D is the changes of each subtype of CD8 T cells in breast lesion samples of the CR cohort before and after treatment, among which the proportion of ZNF683+CD8 T cells increased after treatment (median test, P<0.05); Part E is the changes of each subtype of CD8 T cells in breast lesion samples of the SD cohort before and after treatment; Part F is the changes of each subtype of CD8 T cells in the draining lymph node samples of the CR cohort before and after treatment, among which the proportion of ZNF683+CD8 T cells decreased after treatment while the proportion of IL7R+CD8 T cells increased after treatment (one-sided t test, P<0.05).

[0014] Figure 3 The figure is a graphical representation of the changes in clonal frequency of each CD8 T cell subtype before and after treatment, where Part A is a graph showing the changes in clonal frequency of CD8 T cells in breast lesions of the CR cohort before and after treatment; Part B is a graph showing the changes in clonal frequency of CD8 T cells in breast lesions of the SD cohort before and after treatment; Part C is a graph showing the changes in clonal frequency of CD8 T cells in draining lymph nodes of the CR cohort before and after treatment; Part D is a graph showing the proportion of each CD8 T cell subtype in CD8 T cells with and without amplification in breast lesions of the CR cohort; Part E is a graph showing the proportion of each CD8 T cell subtype in CD8 T cells with and without amplification in breast lesions of the SD cohort; Part F is a graph showing the proportion of each CD8 T cell subtype in CD8 T cells with and without amplification in draining lymph nodes of the CR cohort.

[0015] Figure 4It is a graphic representation comparing the changes of various cell subtypes in the lesions and draining lymph node samples in the CR group. Part A is a distribution diagram of the overlapping TCR clonotypes of ZNF683+CD8 T cells and GZMK+CD8 T cells in the breast lesion cells of the CR cohort; Part B is a diagram of the overlapped number of amplified clonotypes between the subtypes of CD8 T cells in the breast lesion cells of the CR cohort, which can represent the proportion of cell subtypes, and there is a statistical difference between the number of overlapped clonotypes with bold borders and the number of overlapped clonotypes of other clonotypes (using Fisher's exact test); Part C is a distribution diagram of the overlapping TCR clonotypes of ZNF683+CD8 T cells and GZMK+CD8 T cells in the draining lymph node cells of the CR cohort; Part D is a diagram of the overlapped number of unamplified clonotypes between the subtypes of CD8 T cells in the draining lymph node cells of the CR cohort, which can represent the proportion of cell subtypes, and there is a statistical difference between the number of overlapped clonotypes with bold borders and the number of overlapped clonotypes of other clonotypes (using Fisher's exact test); Part E is a diagram of the overlapped number of unamplified clonotypes between the subtypes of CD8 T cells in the draining lymph node cells of the SD cohort. The diagram shows the overlap of the expanded clonal types between the T cell subtypes, which can represent the proportion of the cell subtypes. There is a statistical difference between the overlap of the clonal types with bold borders and the overlap of other clonal types (using Fisher's exact test); F is a diagram showing the source of the expanded clonal types of CD8 T cells in the breast lesions of the CR cohort (two-sided t-test, P=0.000343).

[0016] Figure 5 It is a graphic representation of the correlation between the clonal types of ZNF683+CD8 T cells in breast lesions and draining lymph nodes. Part A shows the TCR clonal types amplified in breast lesions of the CR cohort and the TCR clonal types not amplified in draining lymph nodes of the CR cohort. Red represents the clonal types that overlap in breast lesions and draining lymph nodes, and blue represents the clonal types that are unique to breast lesions. B is the projection of the TCR clonal types that overlap in CD8 T cells in breast lesions and draining lymph nodes of the CR cohort in the distribution of CD8 T cells in breast lesions. C is the projection of the TCR clonal types that overlap in CD8 T cells in breast lesions and draining lymph nodes of the CR cohort in the distribution of CD8 T cells in draining lymph nodes. D is the number of clonal type overlaps between each CD8 T cell subtype in breast lesions and draining lymph nodes of the CR cohort. There is a statistical difference between the number of clonal type overlaps with bold borders and the number of other clonal type overlaps (using Fisher's exact test).

[0017] Figure 6This is a differential gene analysis diagram between expanded (E) and non-expanded (NE) CD8 T cells in breast lesions and draining lymph nodes before and after treatment. Part A in the figure is a differential gene analysis diagram between expanded (E) and non-expanded (NE) CD8 T cells in breast lesions before and after treatment; Part B is a differential gene analysis diagram between expanded (E) and non-expanded (NE) CD8 T cells in draining lymph nodes before and after treatment.

[0018] Figure 7 Figure 3 is a graphic representation of the interaction analysis between endothelial cells and CD8 TRM cells in breast lesion samples from the CR cohort.

[0019] Figure 8 Figure 2 is a graphic representation of the typing of blood CD8 T cells in CR and SD samples.

[0020] In the specific implementation manner, the sub-figures in each figure are represented by: figure + sub-figure number, for example, Figure 1 The sub-figure of part A in the figure is represented by Figure 1 A, Attachment Figure 2 The C part of the sub-figure is represented by Figure 2 C, and so on. DETAILED DESCRIPTION

[0021] The present invention will be further described below in conjunction with specific experiments.

[0022] The idea of ​​this experiment is to use multi-omics methods such as single-cell sequencing, T cell receptor (TCR) sequencing, B cell receptor (BCR) sequencing, combined with experimental methods such as immunofluorescence and immunohistochemistry, to group the CD8+T cells in breast lesion samples and draining lymph node samples in the CR and SD groups of TNBC patients with neoadjuvant chemotherapy, compare the changes in the subpopulations and numbers of CD8+T cells in the two samples of each group before and after treatment, study the effect of NAC on immune cells in the TME and TdLN of TNBC patients, and discover patterns with clinical value.

[0023] 1. Experimental Methods

[0024] 1. Preparation of Breast Lesion and Axillary Lymph Node Cell Suspensions

[0025] 1.1 Tissue cleaning and pretreatment:

[0026] Breast lesions and axillary lymph node samples were collected, and the sample tissues were washed in a 50 mL centrifuge tube with 1× Hanks' balanced salt solution (HBSS) containing 0.04% fetal bovine serum, without Ca2+ / Mg2+, and at a concentration of 400 μg / mL. A small amount of 1× HBSS and 1-2 mL of DMEM culture medium were added, and the sample tissues were stored in a cell culture dish.

[0027] 1.2 Tissue dissociation:

[0028] Cut the tissue into pieces less than 2 mm and resuspend the tissue pieces in 3-5 mL of DMEM medium containing digestion reagent. Transfer the tissue in the culture dish to a 5 mL centrifuge tube and incubate in a 37°C constant temperature water bath for 30-100 min. Gently vortex once every 5 min during the incubation process.

[0029] The digestion reagent was prepared as follows: 2.5 mL DMEM medium, 0.1% 1 mg / mL collagenase I, 0.1% 1 mg / mL neutral protease, 0.02% (0.2 mg / ml) deoxyribonuclease I, and 0.01% hyaluronidase.

[0030] 1.3 Termination of digestion:

[0031] The cell suspension obtained after digestion was filtered into a 50 mL centrifuge tube using a 100 μm nylon mesh, and the centrifuge tube for digestion reaction was washed with 2 mL DMEM culture medium, and the washed DMEM was transferred to the same 50 mL centrifuge tube; the tissue cell suspension was centrifuged at 350 g for 5 min at 4 °C in a centrifuge, and the supernatant was transferred to another centrifuge tube and stored until the entire process was completed.

[0032] 1.4 Red blood cell lysis:

[0033] 1) Resuspend the cell pellet obtained in the previous step with 3-5 mL 1× red blood cell lysis buffer, place on ice for 5 minutes, then add 10 mL pre-cooled Wash Buffer, and centrifuge at 350g for 5 minutes at 4°C;

[0034] 2) Remove the supernatant, resuspend the cells in 10 mL of HBSS, and then pipette up and down 5 times with a wide-bore pipette tip;

[0035] 3) Centrifuge the tissue cell suspension at 350 g for 5 min at 4°C in a centrifuge, and then transfer the supernatant to another centrifuge tube and store it until the end of the entire process;

[0036] Repeat the above resuspension and centrifugation steps, and wash the single cell suspension twice with HBSS to remove cell debris; repeat the above steps no more than 5 times to ensure cell viability; use a cell debris removal kit if necessary.

[0037] 1.5 Microscopic examination and counting:

[0038] 1) Collect the cell supernatant, add 50-1000 μL of pre-cooled HBSS according to the volume, and mix gently;

[0039] 2) Filter the obtained cell suspension into a 50 mL centrifuge tube using a 40 μm nylon mesh, take 1 μL of the cell suspension and stain it with trypan blue to detect cell viability and concentration under an optical microscope;

[0040] 3) The ideal cell concentration is 700-1200 cells / μL. HBSS can be used to dilute the cell suspension if necessary.

[0041] 2. Preparation of peripheral blood samples:

[0042] Peripheral blood samples were collected using EDTA-K2 vacuum tubes for cell extraction to obtain peripheral blood mononuclear cells (PBMCs); the PBMC cell layer was extracted, 5 volumes of PBS were added, centrifuged at 300 g for 10 min at 4°C, and washed twice; the supernatant was removed, and HBSS was added to resuspend the cells until the concentration reached 700-1200 cells / μL.

[0043] 3. scRNA-seq library construction and sequencing:

[0044] The prepared and tested single cell suspension was placed on ice, and a 10× single cell library was constructed and the single cell transcriptome and TCR group were sequenced within 30 minutes.

[0045] 4. Clustering of peripheral blood cells and breast lesion and axillary lymph node infiltrating cells:

[0046] 4.1 Expression matrix normalization

[0047] The obtained UMI count matrix was further processed using Seurat (4.3.0). The expression level of each single-cell gene in peripheral blood, breast lesions, and axillary lymph node infiltration was normalized using the “NormalizationData” function with default parameters.

[0048] 4.2 Identification of highly variable genes

[0049] 2000 highly variable genes were identified using the “FindVariableFeatures” function of the “vst” method. All TRAV, TRBV, TRDV, and TRGV genes were removed from the list of highly variable genes in the T cell clustering analysis to avoid the significant effect of T cell receptor variable genes.

[0050] 4.3 Regulation Expression Matrix

[0051] After removing heteroplasmy associated with mitochondrial contamination, the gene expression matrix was scaled and centered using the “ScaleData” function.

[0052] 4.4 Principal Component Analysis

[0053] The first 20 principal components were used to construct a shared nearest neighbor (SNN) image using the “FindNeighbor” function.

[0054] 4.5 Cell Clustering

[0055] Cell clustering was determined using the Louvain algorithm; resolution was determined based on the principle that each cell population expresses a unique set of genes.

[0056] 4.6 Visualization

[0057] The above SNN images are displayed using Uniform manifold approximation and projection (UMAP) to visualize the single-cell transcriptome profile.

[0058] 4.7 Cell population annotation, data visualization and differential gene analysis

[0059] 1) Cell population annotation

[0060] Basically, well-studied immune cell type marker genes are used to observe whether these genes are among the differentially expressed (DE) genes between the studied cell population and the remaining cell populations to assign the cell population to the most likely identity.

[0061] 2) Differential gene analysis

[0062] Use the "FindAllMarkers" function in Seurat and set the parameters to min.pct = 0.25 and min.log2FC = 0.25. In this way, the DE genes in each cell population or cell subpopulation can be identified. When non-immune cell highly expressed genes appear in the DE genes, they are manually removed.

[0063] 3) CD8 T cell recognition

[0064] Since there is a lot of transcriptome overlap between NK cells and CD8 T cells, it is necessary to first distinguish NK cells from CD8 T cells in the process of identifying CD8 T cell populations, and it is necessary to use T cell marker genes (CD3D, CD3E, and CD3G) and NK cell marker genes (CD56, NKG7). In this study, cells with high expression of NK cell marker genes NKG7 and CD56 were eliminated from the CD8T cell subtype.

[0065] 5. Pseudo-time analysis

[0066] Slingshot (v2.6.0, R software package) was used to perform pseudo-time series analysis. The “UMAP” matrix was used in the “slingshot” function. Since CCR7 is known to represent a naive cell-related phenotype, “CCR7+CD8 T cells” were used as the starting point of the differentiation trajectory.

[0067] 6. Calculation of gene set activity scores

[0068] AUCell (v1.20.2, R software package) was used to predict the activity score of the corresponding gene set in each single cell based on the ranking of all genes in each single cell, and AUCell was used to establish the area under the recovery curve. The X-axis shows the ranking of highly expressed genes according to expression values, and the Y-axis shows the number of genes in the analyzed gene set. Gene activity scores can be obtained by measuring the proportion of expressed genes in the studied gene set and comparing the expression levels of the studied gene set and other genes in each cell. The CD8TRM gene series includes cytotoxic gene sets (PRF1, GZMB, GZMA, GZMH, NKG7, GNLY), inhibitory signal gene sets (PDCD1, CTLA4, HAVCR2, LAG3, TIGIT), and NK cell gene sets (KLRD1, FGFBP2, FCGR3A, S1PR5, KLRC1, KLRC3, KLRB1, KLRC2).

[0069] 7. Functional Enrichment Analysis

[0070] Gene sets of expanded (E) and non-expanded (NE) cells were compared, and fgsea (V1.24.0, R software package) was used to perform rapid signature gene set enrichment analysis. The hallmarker gene set in the Molecular Signature Gene Set Database (MsigDB) was used to download pathway-related gene sets. These gene sets and genes with logarithmic averaged expression fold differences were subjected to fgsea enrichment analysis using the fgsea function (setting parameters as minSize=15, maxSize=500, nperm=1000). Functional enrichment analysis of biological states or processes was performed only for hallmarker gene sets with adjusted P values ​​(adj.P value) <0.05.

[0071] 8. Analysis of interactions between CD8 TRM cells, DCs, endothelial cells, and epithelial cells

[0072] Cellchat (v1.6.1, R package) was used to assess the interactions between breast lesions and draining lymph nodes with T cells, non-immune cells, and DCs. This package allows visualization of cell-to-cell interactions based on known ligand-receptor pairs. Interactions between epithelial cells and DCs were first analyzed; then DC-to-CD8 TRM cell and endothelial-to-CD8 TRM cell interactions were compared before and after treatment. Most analyses were performed using default parameters (except raw.use=TRUE, population.size=TRUE in the “computeCommunProb” function). All datasets from the three Cellchat databases “Cell-to-cell contact”, “Secreted Signaling”, and ECM receptors were used to assess cell-to-cell interactions. First, the samples before and after treatment were analyzed separately, and then the objects before and after treatment were merged and compared using the "mergeCellchat" function. Finally, the "netVisual_bubble" function was applied to visualize the communication probability between various ligand-receptor groups in different cell directions, including epithelial cells acting on DCs, DCs acting on CD8 TRM cells, endothelial cells acting on CD8 TRM cells, DCs acting on epithelial cells, and CD8 TRM cells acting on endothelial cells.

[0073] 9. Single-cell sequencing data processing of T cell receptor (TCR) in blood, breast lesions and draining lymph nodes

[0074] TCR FASTQ reads were processed using the ‘cellranger vdj’ function in Cell Ranger and applying human GRCh38 V(D)J primers to generate single-cell V(D)J sequences and annotate each library. 72.62% (draining lymph nodes) and 82.69% (breast lesions) of the clonotypes contained one TRA and one TRB sequence; 2.80% (draining lymph nodes) and 0.26% (breast lesions) of the clonotypes contained two TRA and two TRB sequences; 3.73% (draining lymph nodes) and 5.46% (breast lesions) of the clonotypes contained one TRA and two TRB sequences; 2.26% (draining lymph nodes) and 1.44% (breast lesions) of the clonotypes contained one TRA sequence; 11.34% (draining lymph nodes) and 9.89% (breast lesions) of the clonotypes contained one TRB sequence; 7.24% (draining lymph nodes) of the clonotypes contained two TRA and one TRB sequence.

[0075] 10. Identify newly generated TCRs and clonal types after treatment

[0076] The frequency of each TCR in the total TCR frequency was compared before and after treatment. If the frequency of a TCR increased after treatment, it was defined as an expanded (E) clonotype, otherwise it was a non-expanded (NE) clonotype. If an expanded clonotype did not exist before treatment, it was defined as an emergent clonotype. Before and after treatment, there were >11,000 and >2,000 valid breast lesion TCRβ sequences, >4,000 and >11,000 valid draining lymph node TCRβ sequences, and >1,000 and >1,000 valid blood TCRβ sequences, respectively.

[0077] 11. Analyze and map the same TCR sequences shared by different T cell populations

[0078] To effectively evaluate the relationship between different T cell populations, the TCR sequences shared by different T cell populations were analyzed. The clonotype information was merged with the scRNA sequencing dataset to map the selected breast lesion cell clonotypes to TdLN and blood samples. The clonotype information of the extracted cell subtypes was used, and the Dimplot function of Seurat (v4.3.0, R software package) was used to present the CD8 T cells in the draining lymph nodes or peripheral blood, so that the extracted breast lesion T cell clonotypes were mapped to the draining lymph nodes and blood CD8 T cells.

[0079] 12. Identification of the transcriptional regulatory network of CD8 TRM cells

[0080] To identify the transcriptional regulatory network of CD8 TRM cells in draining lymph nodes and breast lesions before and after treatment, the R software package (v1.3.1) was used to visualize the key regulatory genes of the CD8 TRM cell population and its different subtypes. First, a portion of cells were randomly selected for identification of transcriptional regulators. Then, the expression level of each transcriptional regulator in the entire sample was analyzed. For each sample, the R software package pheatmap (v1.0.12) was used to identify unique transcriptional regulators in different cell subtypes. In addition, for each transcriptional regulator, the gene network it regulated was analyzed.

[0081] 13. Statistical methods

[0082] SPSS29.0.1.0 (IBM, USA) was used for statistical analysis. For survival analysis, the Gehan-Breslow-Wilcoxon test was used to calculate the P value, and the Mantel-Haenszel test was used to calculate the attributable risk (Hazard ratio). In the process of comparing the changes in the proportion of cell subtypes in samples before and after treatment, one-sided or two-sided T tests and paired T tests were used. When the P value is less than 0.05, it is expressed as *; when the P value is less than 0.01, it is expressed as **.

[0083] All statistical graphs were drawn using ggplot2 (v3.4.2, R software package) and Graphpad Prism 9.0 (GraphPad Software, USA).

[0084] 2. Experimental Results

[0085] 1. The study found that the T cells that actively fight tumors in neoadjuvant chemotherapy for triple-negative breast cancer show effector memory T cell subtypes.

[0086] The proportion of CD8+ tissue-resident memory T cells in draining lymph nodes and breast lesion samples in different treatment response groups before and after treatment The CD8+ T cells in breast lesion samples of the CR (complete remission) group and the SD (stable disease) group were grouped, and the results were as follows Figure 1 , showing paired samples of 2 patients in the CR cohort before and after treatment (e.g. Figure 1 A) and paired samples of 3 patients in the SD cohort before and after treatment (e.g. Figure 1 D) were well merged without significant batch effect. CD8 T cells in both cohorts included ZNF683+CD8 T cells, IL7R+CD8 T cells, GZMK+CD8 T cells, ISG CD8 T cells and CD3D-NKT cells (such as Figure 1 B, Figure 1 E). Since ZNF683 and ITGAE are marker genes for tissue resident memory T cells (TRM), the ZNF683+ cell population represents CD8 TRM cells. This cell population also highly expresses the cytotoxic marker genes GZMB and GNLY as well as the immune checkpoint molecules PDCD1 and CTLA4 (e.g. Figure 1 C, Figure 1 F).

[0087] CD8 T cells in the draining lymph node samples of the CR cohort were analyzed, and the results were as follows Figure 2 As shown, the draining lymph node samples were well merged (e.g. Figure 2A). CD8 T cells in the draining lymph node samples are also composed of ZNF683+CD8 T cells, GZMK+CD8 T cells, IL7R+CD8 T cells, and ISG CD8 T cells (e.g. Figure 2 B). ZNF683+CD8 T cells highly express tissue-resident genes (ITGAE, ZNF683), cytotoxic genes (GZMB, GNLY), and immune checkpoint genes (PDCD1, CTLA4), showing the characteristics of tissue-resident memory T cells (e.g. Figure 2 C). The changes in the proportion of each subtype of CD8 T cells in the total number of cells in the draining lymph node samples and the breast lesion samples of the CR and SD cohorts before and after treatment were analyzed and compared. The results showed that in the CR cohort samples, the proportion of ZNF683+CD8 T cells in the draining lymph node samples before treatment was high and decreased after treatment; on the contrary, the proportion of ZNF683+CD8 T cells in the breast lesions before treatment was low and increased after treatment; both were statistically significant (such as Figure 2 D. Figure 2 E and Figure 2 F). In contrast to the changes in ZNF683+CD8 T cells, the proportion of IL7R+CD8 T cells in the draining lymph node samples increased after treatment, while the proportion in the breast lesion samples decreased. Since IL7R+CD8 T cells have the characteristics of naive T cells, this change suggests that after treatment, a TRM cell population with a cytotoxic effect was established in the breast lesions, while a decrease in the TRM cell population occurred in the draining lymph nodes.

[0088] 2. The study found that T cells amplified by neoadjuvant chemotherapy for triple-negative breast cancer had subclones with similar phenotypes to T cells in tumor-draining lymph nodes.

[0089] The changes in the clonal frequency of each CD8 T cell subtype in breast lesion and draining lymph node samples in the CR cohort and breast lesion samples in the SD cohort before and after treatment were analyzed. Figure 3 As shown in the figure, in the breast lesion samples of the CR cohort, the cell subtypes with higher frequencies of CD8 T cell clonotypes before treatment were mainly IL7R+CD8 T cells; the cell subtypes with higher frequencies of CD8 T cell clonotypes after treatment were mainly ZNF683+CD8 T cells; there was no statistical difference in the clonotype frequencies of ZNF683+CD8 T cells and other CD8 T cell subtypes before treatment; after treatment, the clonotype frequencies of ZNF683+CD8 T cells were higher than those of other CD8 T cell subtypes, and the difference was statistically significant (such as Figure 3 A). In the breast lesion samples of the SD cohort, the clonal frequency distribution of each CD8 T cell subtype before and after treatment was relatively uniform; there was no statistical difference in the clonal frequency between ZNF683+CD8 T cells and other CD8 T cell subtypes before and after treatment (e.g. Figure 3 B). In contrast to the breast lesion samples in the CR cohort, in the draining lymph node samples of the CR cohort, the cell subtypes with higher frequencies of CD8 T cell clonotypes before treatment were mainly ZNF683+CD8 T cells; after treatment, the cell subtypes with higher frequencies of CD8 T cell clonotypes were mainly IL7R+CD8 T cells; before treatment, the frequency of ZNF683+CD8 T cell clonotypes was higher than that of other CD8 T cell subtypes, and the difference was statistically significant; after treatment, there was no statistical difference in the clonotype frequencies of ZNF683+CD8 T cells and other CD8 T cell subtypes in the draining lymph node samples (e.g. Figure 3 C).

[0090] The changes in the frequencies of each CD8 TCR clonotype before and after treatment were compared. The TCR with increased clonotype frequency after treatment was defined as expanded (E) CD8 T cells, while the TCR with unchanged clonotype frequency after treatment was defined as non-expanded (NE) CD8 T cells. The distribution of each cell subtype in the NE group and the E group was compared, and the results showed that in the breast lesion samples of the CR cohort, the expanded CD8 T cell subtype was mainly ZNF683+CD8 T cells; the non-expanded CD8 T cell subtype was mainly IL7R+CD8 T cells; the proportion of ZNF683+CD8 T cells in the E group was significantly higher than that in the NE group, and the difference was statistically significant (such as Figure 3 D). In the breast lesion samples of the SD cohort, amplified and non-amplified TCRs were evenly distributed in each CD8 T cell subtype; there was no significant difference in the proportion of each cell subtype between the E group and the NE group (e.g. Figure 3 E). In contrast to the breast lesion samples in the CR cohort, in the draining lymph node samples in the CR cohort, the expanded CD8 T cell subtype was mainly IL7R+CD8 T cells; while the non-expanded CD8 T cell subtype was mainly ZNF683+CD8 T cells; the proportion of ZNF683+CD8 T cells in the NE group was significantly higher than that in the E group, while the proportion of IL7R+CD8 T cells in the E group was significantly higher than that in the NE group (e.g. Figure 3 F).

[0091] The above analysis results showed that ZNF683+CD8 T cells in breast lesion samples of the CR cohort expanded after treatment, while ZNF683+CD8 T cells in breast lesion samples of the SD cohort did not change significantly; in contrast to the changes in CD8 TRM in breast lesions, the proportion of ZNF683+CD8 T cells in draining lymph node samples of the CR cohort decreased after treatment, suggesting the establishment of CD8+ tissue resident memory T cells (TRM) in breast lesions of TNBC patients after chemotherapy, and the destruction of CD8 TRM cell populations in draining lymph nodes; this change only existed in the CR cohort with good treatment response. Considering that draining lymph nodes serve as a storage depot for delivering lymphocytes to the TME, this change suggests that CD8 TRM delivery from draining lymph nodes to breast lesions may occur in TNBC patients after chemotherapy.

[0092] Comparison of the changes in each cell subtype in the lesion and draining lymph node samples in the CR group showed that Figure 4 The distribution of expanded clonal types in each cell subtype of breast lesion samples in the CR group was compared. Figure 4 A, showing that the cell subtype with the most overlapping clonal types with other cell subtypes is ZNF7683+CD8 T cells (see Figure 4 B). The same method was used to compare the distribution of non-expanded clonal types in each cell subtype of the draining lymph node samples in the CR cohort (e.g. Figure 4 C), the results showed that the cell subtype with the most overlapping clonal types with other cell subtypes was also ZNF683+CD8 T cells (see Figure 4 D). The same method was used to compare the distribution of expanded clonal types in each cell subtype of the breast lesion samples of the SD cohort. The results showed that the cell subtypes with the most overlapping clonal types with other cell subtypes were GZMK+CD8 T cells and ISGCD8 T cells (see Figure 4 E). The results of this analysis suggest that ZNF7683+CD8 T cells are specifically formed in the breast lesion samples of the CR cohort and disappear in the draining lymph node samples of the CR cohort (see Figure 4 F).

[0093] 3. The study found that the number of ZNF683+ effector memory CD8T cells in the tumor during neoadjuvant chemotherapy for triple-negative breast cancer was positively correlated with the efficacy of neoadjuvant chemotherapy.

[0094] Correlation between the clonal types of ZNF683+CD8 T cells in breast lesions and draining lymph nodes Figure 5As shown in Figure 2, the clonal types that just appeared after treatment (Emergent) and the clonal types that existed before treatment and expanded after treatment (Pre-existing) in breast lesions were compared with the clonal types of ZNF683+CD8 T cells in the draining lymph nodes. The results are shown in Figure 2. Figure 5 As shown in A, both the Emergent and Pre-existing clonotypes in breast lesions overlapped with those in draining lymph nodes. The ZNF683+CD8 T cell clonotypes that were expanded (E) in breast lesions were compared with the ZNF683+CD8 T cell clonotypes that were not expanded (NE) in draining lymph nodes, and a total of 134 overlapping clonotypes were found. The 134 overlapping clonotypes were projected onto breast lesion CD8 T cells and draining lymph node CD8 T cells, respectively. The results are shown in Figure 5 B and Figure 5 As shown in C, the results show that the clonal types in breast lesions and draining lymph nodes are enriched in the ZNF683+CD8 T cell subtype. The clonal types of each subtype of CD8 T cells in breast lesions and draining lymph nodes were compared, and the results showed that the number of clonal type overlaps between ZNF683+CD8 T cells in breast lesions and draining lymph nodes was the highest. Figure 5 As shown in D, the difference is statistically significant.

[0095] like Figure 6 As shown, the differentially expressed genes between expanded (E) and non-expanded (NE) CD8 T cells in breast lesions and draining lymph nodes before and after treatment were compared. The results showed that genes related to CD8 T cell activation and effector function, TCR signaling genes, and tissue-resident memory genes were mainly enriched in CD8 T cells expanded (E) in breast lesions before and after treatment and CD8 T cells non-expanded (NE) in draining lymph nodes before and after treatment; while genes related to initial CD8 T cells were mainly enriched in CD8 T cells non-expanded (NE) in breast lesions before and after treatment and CD8 T cells expanded (E) in draining lymph nodes before and after treatment.

[0096] To clarify which cell subtypes of CD8 TRM cells were most strongly recruited by endothelial cells in breast lesion samples of the CR cohort after treatment, the interaction between endothelial cells and CD8 TRM cells in breast lesion samples of the CR cohort was analyzed. Figure 7 As shown, the interaction between endothelial cells and CD8 TRM cells was most significantly enriched in the proliferating CD8 TRM cell and stem-like CD8 TRM cell subtypes.

[0097] Since immune cells in the blood are the first to be affected by chemotherapy, analyzing the changes in CD8 T cells in the blood before and after treatment is helpful to explore the effects of chemotherapy on promoting systemic anti-tumor immunity. Figure 8 To this end, this experiment analyzed the typing of blood CD8 T cells in CR and SD samples, and the results are shown in Figure 8 A. Figure 8 B and Figure 8 As shown in Figure C, blood CD8 T cells are divided into eight subtypes: ZNF683+CD8 T cells with characteristics of colonizing memory T cells, GZMK+CD8 T cells, resting NK T cells, CCR7+CD8 T cells, cytotoxic NK T cells, Vγ9Vδ2Tγδ cells, MKI67+CD8 T cells, ISG CD8 T cells and GZMB+CD8 T cells.

[0098] It is worth noting that the TCRs of all CD8 T cells in breast lesion samples of the SD group and the CR group were projected onto the UMAP images of CD8 T cells in the blood samples of the corresponding groups. The results showed that compared with the SD group, the TCRs of CD8 T cells in breast lesion samples of the CR group were more significantly enriched in ZNF683+CD8 T cells in the blood CD8 T cell samples (such as Figure 8 D). The same method was used to project the TCR of CD8 T cells amplified in breast lesion samples of the SD group and the CR group onto the UMAP images of CD8 T cells in the blood samples of the corresponding groups. The results also showed that compared with the SD group, the TCR of CD8 T cells in breast lesion samples of the CR group was more significantly enriched in ZNF683+CD8 T cells in the blood CD8 T cell samples (as shown in Figure 8 E), suggesting that the level of effector memory T cells ZNF683+CD8 T cells in the blood of TNBC patients before and after treatment may become a biomarker for predicting patients' response to chemotherapy.

Claims

1. ZNF683 + Effector memory CD8 T cells as a biomarker for predicting chemotherapy efficacy in triple-negative breast cancer.

2. ZNF683 + Use of effector memory CD8 T cells in the preparation of a kit for predicting the efficacy of chemotherapy for triple-negative breast cancer.

3. The use according to claim 2, characterized in that: The kit includes a method for ZNF683 + Effector memory CD8 T cells are detected qualitatively and / or quantitatively.

4. The use according to claim 3, characterized in that: The substances for qualitative and / or quantitative detection include substances for identifying the cell marker gene ZNF683.

5. The use according to claim 1, characterized in that: The use includes identifying ZNF683 by single cell sequencing + Effector memory CD8 T cell approach.

6. The use according to claim 1, characterized in that: ZNF683 + Effector memory CD8 T cells were obtained from the blood of patients with triple-negative breast cancer.

Images