Application of a combination of diagnostic biomarkers for lung adenocarcinoma brain metastases based on BNIP3, CD74, and LYZ

Abstract

This invention relates to the field of biomedical technology, providing the application of a combination of diagnostic biomarkers for lung adenocarcinoma brain metastases based on BNIP3, CD74, and LYZ. Specifically, it relates to the application of this combination of diagnostic biomarkers in the preparation of reagents for predicting the prognosis of lung adenocarcinoma brain metastases. These reagents predict the prognosis of lung adenocarcinoma brain metastases by detecting the expression levels of BNIP3, CD74, and LYZ in the brain metastatic tissues of patients with lung adenocarcinoma brain metastases. This invention, based on a combination of diagnostic biomarkers for lung adenocarcinoma brain metastases based on BNIP3, CD74, and LYZ, can accurately assess the progression and risk level of a patient's disease in predicting prognosis and risk stratification of lung adenocarcinoma brain metastases. This provides precise reference for subsequent individualized treatment plans and rehabilitation programs, thus optimizing patient management.

Description

Technical Field

[0001] This invention belongs to the field of biomedical technology, and in particular relates to the application of a combination of diagnostic biomarkers for lung adenocarcinoma brain metastasis based on BNIP3, CD74 and LYZ. Background Technology

[0002] Lung adenocarcinoma (LUAD) is one of the most common types of lung cancer, and brain metastasis (BM) is a significant factor contributing to poor prognosis. Existing research indicates that the incidence of brain metastasis in lung adenocarcinoma patients is gradually increasing, especially in advanced-stage patients. Although several biomarkers have been proposed for the diagnosis and prognostic assessment of lung adenocarcinoma, specific biomarkers for brain metastasis remain scarce. Current detection methods largely rely on imaging examinations and clinical symptoms, lacking effective means for early identification of high-risk patients.

[0003] Currently, the diagnosis of lung adenocarcinoma mainly relies on tissue biopsy and imaging examinations such as CT and MRI. While these methods provide some diagnostic information, they often fail to accurately identify the risk of brain metastasis in the early stages. Furthermore, existing biomarkers, such as EGFR mutations and ALK rearrangements, while important in the treatment of lung adenocarcinoma, have limited predictive ability for brain metastasis. With the development of molecular biology techniques, more and more research is focusing on the application of circulating tumor DNA and other biomarkers in cancer diagnosis.

[0004] Despite the progress made in the diagnosis and treatment of lung adenocarcinoma by existing technologies, the following major problems still exist: (1) Difficulty in early identification: Existing diagnostic methods often only detect brain metastases after they occur, causing patients to miss the best treatment opportunity; (2) Lack of specificity of biomarkers: Most existing biomarkers are general and lack specificity for brain metastases, making it impossible to effectively distinguish high-risk patients; (3) Limitations in clinical application: Most existing detection methods are invasive, have low patient acceptance, and provide insufficient support for clinical decision-making; (4) Lack of evidence for targeted therapy: Due to the lack of effective biomarkers, targeted therapy strategies for brain metastases have not been fully developed, affecting patients' survival rate and quality of life.

[0005] Therefore, in-depth exploration of the molecular mechanisms of brain metastasis in lung adenocarcinoma and the search for new biomarkers are of great significance for early identification of high-risk brain metastasis patients and optimization of clinical decision-making. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an application of a combination of diagnostic biomarkers for lung adenocarcinoma brain metastases based on BNIP3, CD74, and LYZ, with the aim of resolving the problems mentioned in the background art.

[0007] In a first aspect, the present invention provides the application of a combination of diagnostic biomarkers for lung adenocarcinoma brain metastasis based on BNIP3, CD74 and LYZ in the preparation of a reagent for predicting the prognosis of lung adenocarcinoma brain metastasis. The reagent is used to predict the prognosis of lung adenocarcinoma brain metastasis by detecting the expression levels of BNIP3, CD74 and LYZ in the brain metastasis tissue of patients with lung adenocarcinoma brain metastasis.

[0008] Furthermore, the reagents include those for detecting the protein expression levels and / or mRNA expression levels of BNIP3, CD74, and LYZ.

[0009] Furthermore, the reagents for detecting the protein expression levels of BNIP3, CD74, and LYZ include anti-BNIP3 antibody, anti-CD74 antibody, and anti-LYZ antibody for immunohistochemistry, immunofluorescence, or Western blotting.

[0010] Furthermore, the reagents for detecting the mRNA expression levels of BNIP3, CD74, and LYZ include PCR primers for the cDNA strands of the BNIP3, CD74, and LYZ genes used for qRT-PCR, RNA sequencing, or fluorescence in situ hybridization.

[0011] Secondly, the present invention provides the application of an antibody for detecting the BNIP3+CD74-LYZ-epithelial cell subset in the preparation of a diagnostic kit for lung adenocarcinoma brain metastasis. The kit diagnoses whether a lung adenocarcinoma patient has developed brain metastasis by detecting the proportion of the BNIP3+CD74-LYZ-epithelial cell subset in the primary tumor tissue of the lung adenocarcinoma patient.

[0012] Furthermore, antibodies against the BNIP3+CD74-LYZ- epithelial cell subset include anti-BNIP3 antibodies, anti-CD74 antibodies, and anti-LYZ antibodies for immunohistochemistry or immunofluorescence.

[0013] Furthermore, the proportion of BNIP3+CD74-LYZ- epithelial cell subsets was increased in patients with lung adenocarcinoma brain metastases.

[0014] Furthermore, the lung adenocarcinoma patients include suspected lung adenocarcinoma patients, confirmed lung adenocarcinoma patients, suspected lung adenocarcinoma brain metastasis patients or confirmed lung adenocarcinoma brain metastasis patients, schwannomas patients or glioma patients.

[0015] The present invention has the following beneficial effects:

[0016] (1) A combination of diagnostic markers for brain metastasis of lung adenocarcinoma based on BNIP3, CD74 and LYZ can be applied to the prognosis of brain metastasis of lung adenocarcinoma. In the prognosis prediction and risk stratification of brain metastasis of lung adenocarcinoma, it can accurately assess the disease development trend and risk level of patients, provide accurate reference for subsequent individualized treatment plans and rehabilitation plans, and optimize patient management.

[0017] (2) By detecting the proportion of BNIP3+CD74-LYZ- epithelial cell subsets in the primary tumor tissue of lung adenocarcinoma patients, the brain metastasis of lung adenocarcinoma patients can be diagnosed. It was also found that the BNIP3+CD74-LYZ- epithelial cell subsets interact with immune cells through the MIF signaling pathway to regulate tumor metastasis. This method does not rely on complex methods such as brain imaging, and can detect the risk of brain metastasis earlier, helping doctors to formulate intervention plans in advance to delay disease progression. At the same time, it provides a precise target reference for subsequent research on targeted regulation of the MIF signaling pathway and blocking of metastasis. Attached Figure Description

[0018] Exemplary embodiments of the present invention can be more fully understood by referring to the following figures:

[0019] Figure 1 This is the identification of a pivotal gene related to brain metastasis of lung adenocarcinoma in Example 1 of the present invention, wherein:

[0020] Figure 1 In the diagram, A represents the UMAP plot that divides the primary lung adenocarcinoma tumor and brain metastases into 11 clusters.

[0021] Figure 1 In the diagram, B represents the UMAP images showing the main cell types of the primary lung adenocarcinoma tumor and brain metastases.

[0022] Figure 1 In this context, C represents the fusion dynamic gene dendrogram, which displays clusters of all genes.

[0023] Figure 1 In the diagram, D stands for Module-Typology Heatmap, which illustrates the association between gene modules and brain metastasis.

[0024] Figure 1 E in the graph represents a heatmap of differentially expressed genes from the GSE248830 dataset.

[0025] Figure 1 F in the Venn diagram represents the intersection of differentially expressed genes in single cells, genes from the WGCNA module, and differentially expressed genes from batch RNA sequencing.

[0026] Figure 1 In the figure, G represents the five intersecting genes in the GSE248830 dataset labeled with the volcano icon.

[0027] Figure 2This is the molecular subtyping and pathway enrichment analysis of lung adenocarcinoma brain metastases in Example 2 of the present invention, wherein:

[0028] Figure 2 In this context, A is the consensus clustering matrix for k=2;

[0029] Figure 2 In this context, B is the area under curve C for k=2-9;

[0030] Figure 2 C in the equation represents the consensus distribution curve for k=2-9.

[0031] Figure 2 In this context, D represents the total survival curves of the two clusters.

[0032] Figure 2 In this context, E represents the progression-free survival curves for the two clusters.

[0033] Figure 2 F in the figure represents a box plot showing the differential expression of brain metastasis-related genes between the two clusters.

[0034] Figure 2 G in the diagram represents the KEGG pathway GSEA enrichment map of two clusters of highly expressed genes.

[0035] Figure 3 This is an analysis of the TIME (tumor immune microenvironment), TMB (tumor mutation burden), and chemosensitivity of two clusters in Example 3 of the present invention, wherein:

[0036] Figure 3 In the figure, A is a box plot showing the differences in matrix, immunity, and ESTIMATE scores;

[0037] Figure 3 In this context, B represents the difference in immune infiltration as assessed by CIBERSORT.

[0038] Figure 3 In this context, C represents the difference in immune infiltration assessed by ssGSEA;

[0039] Figure 3 In this context, D represents the difference in tumor mutation burden.

[0040] Figure 3 E in the figure is a waterfall plot of the top 15 mutated genes in the two clusters;

[0041] Figure 3 F in the figure represents the overall survival curves grouped according to different clusters and tumor mutation burden levels.

[0042] Figure 3 G in the figure represents a comparison of the efficacy of 12 chemotherapy drugs.

[0043] Figure 4This is the intercellular communication network between epithelial cells expressing a specific hub gene and other subpopulations in brain metastasis according to Embodiment 4 of the present invention, wherein:

[0044] Figure 4 A in the diagram is a circular plot that shows in detail the interaction strength between epithelial cell subsets that express BNIP3 and those that do not express BNIP3 and other cell subsets in brain metastases.

[0045] Figure 4 B in the diagram is a circular plot that shows in detail the interaction strength between epithelial cell subsets that express CD74 and those that do not express CD74 and other cell subsets in brain metastases.

[0046] Figure 4 C in the diagram is a ring diagram that shows in detail the interaction strength between epithelial cell subsets that express LYZ and those that do not express LYZ and other cell subsets in brain metastases;

[0047] Figure 4 D in the diagram is a ring diagram that shows in detail the interaction strength between epithelial cell subsets expressing BNIP3 but not CD74 or LYZ and other epithelial cell subsets and other cell subsets in brain metastases.

[0048] Figure 4 E in the dot plot highlights the main signaling pathway between BNIP3+CD74-LYZ-epithelial cells and other cell subsets in brain metastases;

[0049] Figure 4 The F in the scatter plot (left) visualizes all pathways and the main sending and receiving cells of the MIF pathway, while the chord plot (right) shows the significant interactions and communication probabilities of the MIF pathway.

[0050] Figure 4 In the diagram, G represents the two ligand-receptor pairs in the MIF pathway. The circular diagram (left) shows the interaction strength, and the violin diagram (right) shows the gene expression levels of the ligand and receptor.

[0051] Figure 5 This is a circular diagram showing the interaction strength between positive and negative expression of IRF1 and TYMP in brain metastases and other cell subsets in Example 4 of the present invention, wherein:

[0052] Figure 5 A in this context is IRF1;

[0053] Figure 5 B in the text stands for TYMP.

[0054] Figure 6 This is the communication analysis of BNIP3+CD74-LYZ-epithelial cells and T cell subsets in Example 5 of the present invention, wherein:

[0055] Figure 6In this context, A represents the re-clustering of T cells into four subgroups;

[0056] Figure 6 B in the diagram represents the interaction strength between epithelial cells and the four T cell subsets.

[0057] Figure 6 In the diagram, C represents the dot plot illustrating the main signaling pathways between epithelial cells and T cell subsets;

[0058] Figure 6 In the diagram, D represents the contribution of a specific T cell subset to the output / input signals of epithelial cells, as shown in the heatmap.

[0059] Figure 6 In the diagram, E represents the chord diagram showing significant interactions and communication probabilities within the MIF pathway;

[0060] Figure 6 In the heatmap, F represents the significant interactions and communication probabilities of the MIF pathway.

[0061] Figure 6 G in the diagram represents the circular diagram of ligand-receptor pairs in the MIF pathway and the violin diagram of gene expression.

[0062] Figure 7 This is the communication analysis of BNIP3+CD74-LYZ-epithelial cells and B cell subsets in Example 5 of the present invention, wherein:

[0063] Figure 7 In this case, A represents B cells that have been re-clustered into three subgroups;

[0064] Figure 7 In the diagram, B represents the interaction strength between epithelial cells and B cell subsets.

[0065] Figure 7 In the diagram, C represents the dot plot illustrating the main signaling pathways between epithelial cells and B cell subsets.

[0066] Figure 7 In the diagram, D represents the contribution of B cell subsets to the output / input signals of epithelial cells, as shown in the heatmap.

[0067] Figure 7 In the diagram, E represents the chord diagram showing significant interactions and communication probabilities within the MIF pathway;

[0068] Figure 7 In the heatmap, F represents the significant interactions and communication probabilities of the MIF pathway.

[0069] Figure 7 G in the diagram represents the circular diagram of ligand-receptor pairs in the MIF pathway and the violin diagram of gene expression.

[0070] Figure 8This is a communication analysis of BNIP3+CD74-LYZ- epithelial cells and myeloid cell subsets in Example 5 of the present invention, wherein:

[0071] Figure 8 In this context, A represents the re-clustering of myeloid cells into 5 subgroups;

[0072] Figure 8 B in the diagram represents the interaction strength between epithelial cells and myeloid cell subsets.

[0073] Figure 8 In the diagram, C represents the dot plot illustrating the main signaling pathways between epithelial cells and myeloid cell subsets.

[0074] Figure 8 In the heatmap, D represents the contribution of myeloid cell subsets to the input / output signals of epithelial cells.

[0075] Figure 8 In the diagram, E represents the chord diagram showing significant interactions and communication probabilities within the MIF pathway;

[0076] Figure 8 In the heatmap, F represents the significant interactions and communication probabilities of the MIF pathway.

[0077] Figure 8 G in the diagram represents the circular diagram of ligand-receptor pairs in the MIF pathway and the violin diagram of gene expression.

[0078] Figure 9 This is an analysis of the distribution and functional enrichment of BNIP3+CD74-LYZ- epithelial cells in epithelial subsets in Example 6 of the present invention, wherein:

[0079] Figure 9 In the diagram, A represents the UMAP diagram, which divides epithelial cells into 6 subgroups.

[0080] Figure 9 In the diagram, B represents the dot plot showing the expression of marker genes in each epithelial subgroup.

[0081] Figure 9 C in the diagram represents the visualization of BNIP3, CD74, and LYZ expression, as well as the distribution of BNIP3+CD74-LYZ-epithelial cells.

[0082] Figure 9 In the diagram, D represents the changes in the proportion of different endothelial cell subsets in positive and negative epithelial cells.

[0083] Figure 9 E in the equation represents the KEGG enrichment analysis of differentially expressed genes between clusters 1 and 4 and clusters 0, 2, 3, and 5.

[0084] Figure 9F in the equation represents the GO enrichment analysis of differentially expressed genes between clusters 1 and 4 and clusters 0, 2, 3, and 5.

[0085] Figure 10 This is the epithelial cell gene set scoring and visualization of Example 7 of the present invention, wherein:

[0086] Figure 10 In the diagram, A represents the enrichment fraction of the migrating gene set in each epithelial subpopulation, as shown by the violin plot, bubble plot, and density scatter plot.

[0087] Figure 10 B in the diagram represents the enrichment fraction of the proliferating gene set in each epithelial subpopulation, displayed by violin plot, bubble plot, and density scatter plot.

[0088] Figure 10 The C in the diagram represents a violin plot illustrating the relationship between BNIP3, CD74, and LYZ expression and proliferation and migration scores.

[0089] Figure 10 In the diagram, D represents the enrichment score of specific gene sets in each epithelial subgroup as shown in the heatmap.

[0090] Figure 11 The Kaplan-Meier survival curves of Example 7 of this invention show the correlation between BNIP3, CD74, and LYZ expression and overall survival and progression-free survival, wherein:

[0091] Figure 11 A in this context is BNIP3;

[0092] Figure 11 B in the text is CD74;

[0093] Figure 11 The C in this context stands for LYZ.

[0094] Figure 12 This is the brain metastasis-related epithelial cell differentiation trajectory analysis of Embodiment 8 of the present invention, wherein:

[0095] Figure 12 In this context, A represents the pseudo-time trajectory of epithelial cells in brain metastases.

[0096] Figure 12 B in the diagram represents the cell density distribution over time.

[0097] Figure 12 In the figure, C represents the expression changes of BNIP3, CD74, and LYZ genes during the pseudo-time series.

[0098] Figure 12 D in the figure is a heatmap of branch-dependent genes in three differentiation states identified through branch expression analysis modeling.

[0099] Figure 12 E in the figure represents the GO enrichment analysis of branch-dependent genes under two different differentiation directions;

[0100] Figure 12 F in the figure represents the KEGG enrichment analysis of branch-dependent genes under two different differentiation directions.

[0101] Figure 13 This invention relates to the tissue expression of BNIP3, CD74, and LYZ in Example 9 and their prognostic correlation in clinical specimens, wherein:

[0102] Figure 13 In Figure A, there are immunohistochemical staining images of brain metastases and paired primary lung adenocarcinoma tumor tissues from two patients with lung adenocarcinoma. Scale bar: 100 μm.

[0103] Figure 13 B in the figure represents the IHC staining of non-metastatic primary lung adenocarcinoma tissue, adjacent normal tissue, schwannoma tissue, and glioma tissue. Scale bar: 100 μm;

[0104] Figure 13 C in the image represents multiplex immunofluorescence staining images of brain metastases, primary lung adenocarcinoma tissue, non-metastatic primary lung adenocarcinoma tissue, and adjacent normal tissue. Scale bar: 50 μm.

[0105] Figure 13 D in the figure represents the Kaplan-Meier curve, which shows the correlation between the expression of BNIP3, CD74, LYZ and their combinations and progression-free survival and overall survival in patients with lung adenocarcinoma.

[0106] Figure 14 This is a statistical chart comparing brain metastasis tissue, primary lung adenocarcinoma tissue without metastasis, adjacent normal tissue, schwannoma tissue, and glioma tissue in Example 9 of the present invention.

[0107] Figure 15 This is the ROC curve of Embodiment 10 of the present invention. Detailed Implementation

[0108] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0109] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention.

[0110] This invention provides the application of a combination of diagnostic biomarkers for lung adenocarcinoma brain metastasis based on BNIP3, CD74, and LYZ in the preparation of a reagent for predicting the prognosis of lung adenocarcinoma brain metastasis. The reagent is used to predict the prognosis of lung adenocarcinoma brain metastasis by detecting the expression levels of BNIP3, CD74, and LYZ in the brain metastasis tissue of patients with lung adenocarcinoma brain metastasis.

[0111] In some embodiments, the reagents include those for detecting the protein expression levels and / or mRNA expression levels of BNIP3, CD74, and LYZ.

[0112] In some embodiments, reagents for detecting the protein expression levels of BNIP3, CD74, and LYZ include anti-BNIP3 antibodies, anti-CD74 antibodies, and anti-LYZ antibodies for immunohistochemistry, immunofluorescence, or Western blotting.

[0113] In some embodiments, the reagents for detecting the mRNA expression levels of BNIP3, CD74, and LYZ include PCR primers for the cDNA strands of the BNIP3, CD74, and LYZ genes for qRT-PCR, RNA sequencing, or fluorescence in situ hybridization.

[0114] In some embodiments, this invention provides the application of an antibody for detecting the BNIP3+CD74-LYZ-epithelial cell subset in the preparation of a diagnostic kit for lung adenocarcinoma brain metastasis. The kit diagnoses whether a lung adenocarcinoma patient has developed brain metastasis by detecting the proportion of the BNIP3+CD74-LYZ-epithelial cell subset in the primary tumor tissue of the lung adenocarcinoma patient.

[0115] In some embodiments, antibodies against the BNIP3+CD74-LYZ-epithelial cell subset include anti-BNIP3 antibodies, anti-CD74 antibodies, and anti-LYZ antibodies for immunohistochemistry or immunofluorescence.

[0116] In some embodiments, the proportion of the BNIP3+CD74-LYZ- epithelial cell subset is increased in patients with lung adenocarcinoma brain metastases.

[0117] In some embodiments, lung adenocarcinoma patients include suspected lung adenocarcinoma patients, confirmed lung adenocarcinoma patients, suspected lung adenocarcinoma brain metastasis patients or confirmed lung adenocarcinoma brain metastasis patients, schwannomas patients or glioma patients.

[0118] Materials and Methods:

[0119] Unless otherwise specified, all bioinformatics methods used below are conventional methods; all materials and reagents used in the following experiments are commercially available unless otherwise specified.

[0120] 1. Data source:

[0121] Single-cell data were obtained from 11 primary lung adenocarcinoma (LUAD PT) and 10 brain metastasis (BM) samples in the GEO dataset GSE131907, and bulk sequencing data were obtained from the GEO dataset GSE248830. In addition, a retrospective analysis was conducted on brain metastasis tissue (BM) from 54 patients with lung adenocarcinoma brain metastases (LUAD BM) at the First Affiliated Hospital of Nanchang University. This included paraffin sections of primary lung adenocarcinoma (PT) from 9 patients, primary non-metastatic lung adenocarcinoma (NMPT) and paired adjacent normal tissue (NAT) from 19 patients with lung adenocarcinoma without brain metastases, paraffin sections of 18 schwannomas, and 19 gliomas.

[0122] 2. Single-cell data processing workflow:

[0123] After screening for cells with a mitochondrial gene proportion of less than 5%, and with 200 to 2500 genes detected and expressed in at least three cells, 33,848 cells from primary lung adenocarcinoma and 16,740 cells from lung adenocarcinoma brain metastases were retained. Subsequently, principal component analysis (PCA) was performed on the top 2000 genes with the largest variations using the R package "Seurat" (version 5.0.1). The Uniform Manifold Approximation and Projection (UMAP) algorithm was used for data visualization in two-dimensional space. Cell annotation was manually performed using cell type markers from previous literature.

[0124] 3. WGCNA analysis (Weighted Gene Co-expression Network Analysis):

[0125] We used the R package "WGCNA" to perform WGCNA analysis on the GEO dataset GSE248830, using primary lung adenocarcinoma and brain metastases as features. By setting the optimal soft threshold, we identified the modules most closely related to primary lung adenocarcinoma and brain metastases, respectively. Finally, we extracted the genes of these modules for subsequent analysis.

[0126] 4. Consistent cluster analysis:

[0127] Based on the expression of characteristic genes of lung adenocarcinoma brain metastases, consensus clustering analysis was performed on GSE248830 using the R package "ConsensusClusterPlus". The "pam" clustering algorithm and "euclidean" distance metric were used to plot a consensus matrix heatmap, a consensus cumulative distribution function (CDF) plot, and an incremental area plot to determine the optimal k value, thereby identifying different lung adenocarcinoma brain metastasis-related clusters.

[0128] 5. GSEA (Gene Set Enrichment Analysis) and Functional Enrichment Analysis:

[0129] Differentially expressed genes (DEGs) were identified using the R package "limma" based on the criteria of |log2FoldChange(FC)| > 1 and adjusted P < 0.05. Subsequently, the entrezIDs of the DEGs were extracted using the R package "org.Hs.eg.db". Functional enrichment analysis of GO and KEGG was performed based on the identified DEGs and entrezIDs using the R packages "clusterProfifiler", "GOplot", "ggplot2", and "enrichplot". Furthermore, the MsigDB gene set was used as an annotation gene set for clustering pathways. Differences in biological signaling pathways between the two groups were explored using GSEA. Enriched pathways were sorted according to consistency scores after screening for P < 0.05.

[0130] 6. TIME (tumor immune microenvironment), TMB (tumor mutation burden), and drug sensitivity analysis:

[0131] The R package "ESTIMATE" was used to calculate the immune and matrix scores for two clusters. The ESTIMATE score, derived by adding the immune and matrix scores, effectively represents a comprehensive assessment of these two scores. Simultaneously, CIBERSORT analysis was used to identify cell types in each sample, aiming to quantify the proportion of 21 different immune cell types in each patient, and ssGSEA was used to compare the abundance of immune function in the two clusters. All patients were divided into four groups based on their comprehensive mutation and risk score levels: high TMB + high risk, low TMB + high risk, high TMB + low risk, and low TMB + low risk. Kaplan-Meier survival curves were used to assess the survival prognosis of the four groups. Subsequently, the R package "maftools" was used to generate waterfall plots to visualize the mutation landscape. The TIDE score was calculated from http: / / tide.dfci.harvard.edu / . Finally, the pRRophetic algorithm was used to evaluate the half-inhibitory concentrations (WICs) of 198 anticancer drugs in LUAD.

[0132] 7. Cell communication analysis:

[0133] Cell-cell communication networks were inferred and visualized using the R package "CellChat" based on molecular interactions provided by the human CellChatDB database. Based on the inferred cell-cell communication networks, communication probabilities were calculated at the cellular, pathway, and gene levels. The results were visualized as circle plots (netVisual_aggregate), chord plots (netVisual_aggregate, netVisual_chord_cell), heatmaps (netVisual_heatmap, netAnalysis_signalingRole_heatmap), bubble plots (netVisual_bubble), violin plots (plotGeneExpression), and scatter plots (netAnalysis_signalingRole_scatter).

[0134] 8. Single-cell gene set score:

[0135] Based on a custom set of proliferation genes (MKI67, IGF1, ITGB2, PDGFC, JAG1, PHGDH) and a set of metastasis genes (HMGB3, MYC, CDK6, CXCL2, MMP7, CXCL8, CCND1, CD44, ITGA3, CDH1) defined in this study, the proliferation and metastasis functions were scored using the AddModuleScore function. The scoring results were visualized using the VlnPlot, DotPlot, and ggplot functions. Furthermore, the scores of gene sets in the MSigDB database were calculated using the GSVA function, and the results were visualized using the R package "pheatmap".

[0136] 9. Pseudo-time series analysis:

[0137] The R package "monocle" (v2.32.0) was used to construct cell differentiation trajectories. Epithelial cells were selected for pseudo-temporal analysis. After selecting genes to define the trajectory, the expression data were projected into a low-dimensional space. Finally, the trajectory plot was drawn using the "plot_cell_trajectory" function, and the cell density plot was drawn using the "ggplot" and "geom_density" functions. The expression changes of marker genes with cell state were visualized using the "plot_genes_jitter", "plot_genes_violin", and "plot_genes_in_pseudotime" functions.

[0138] 10. Immunohistochemistry and immunofluorescence:

[0139] Paraffin sections were cut into 4 μm thin slices, dewaxed and hydrated, and then subjected to antigen retrieval with EDTA buffer. After washing, the slices were incubated with 3% hydrogen peroxide. The slices were blocked with 5% goat serum, then incubated overnight at 4°C with primary antibodies [anti-BNIP3 (GB111204, Servicebio, 1:500), anti-CD74 (GB151179, Servicebio, 1:500), anti-LYZ (15013-1-AP, Proteintech, 1:200)], followed by washing with phosphate buffer. Finally, secondary antibodies were added for labeling and colorimetric reaction, and protein localization and expression were observed using a pathological slide scanner. Two independent pathologists were invited to score the immunohistochemical staining results in a blinded manner.

[0140] Immunohistochemical scoring: Cell staining intensity was graded on a 4-point scale, with no positive staining (negative) scoring 0 points, pale yellow (weakly positive) scoring 1 point, brownish-yellow (positive) scoring 2 points, and brownish-red (strongly positive) scoring 3 points. The percentage of positive cells was also graded on a 4-point scale, with ≤25% scoring 1 point, 26%-50% scoring 2 points, 51%-75% scoring 3 points, and >75% scoring 4 points. The two scores were multiplied to obtain the final score. Three non-repeated fields of view were used for scoring each slide, and the average value was taken. 4μm slides were stained with primary antibody (same as immunohistochemistry) and fluorescently conjugated secondary antibody for immunofluorescence co-localization staining. Cell nuclei were then stained with DAPI (G1012, Servicebio), and the slides were mounted. Finally, the slides were observed under a fluorescence microscope.

[0141] 11. Survival Analysis:

[0142] The overall survival of patients in each group was calculated using the R package "survival", and the survival curves were then visualized using the R package "survminer" through Kaplan-Meier plots.

[0143] 12. Statistical Analysis:

[0144] Data processing was performed using R software (v4.4.1) and related packages. For continuous variables, independent samples t-tests were used to assess differences between groups; for categorical variables, χ² tests were used.

[0145] Example 1: Identification of genes associated with brain metastasis in lung adenocarcinoma

[0146] Eleven primary lung adenocarcinoma tumors and ten brain metastasis samples from the public single-cell dataset GSE131907 were selected for analysis. The results are as follows: Figure 1 As shown.

[0147] After data quality control and standardization, the data was subjected to dimensionality reduction and clustering, resulting in 11 cell clusters: C0-C10 ( Figure 1 Based on classic cell type-specific molecular markers, cells in primary lung adenocarcinoma tumors and brain metastases were annotated into nine cell types: B cells, endothelial cells, epithelial cells, fibroblasts, mast cells, myeloid cells, NK cells, oligodendrocytes, and T cells. Figure 1 (B in the text); then, through WGCNA analysis, the transcriptome data GSE248830 was clustered into 6 modules ( Figure 1 In the C clustering results, the “MEturquoise” module was significantly associated with brain metastasis (p=0.02, r=0.5), and the “MEblue” module was significantly associated with brain metastasis (p=0.008, r=0.55). Figure 1 (D in the text); Simultaneously, differential analysis was performed on the primary lung adenocarcinoma and brain metastasis samples of GSE248830, and the resulting heatmap is shown in Figure 1. Figure 1 As shown in E;

[0148] Since the cancer cells in the primary lung adenocarcinoma tumor and brain metastasis samples are epithelial cells, differential analysis was also performed on the epithelial cells of the single-cell dataset GSE131907. Finally, the intersection of genes identified by three methods—scRNA_epithelial_DEGs, GSE248830_DEGs, and WGCNA analysis—was used to obtain one upregulated (BNIP3) and four downregulated (CD74, LYZ, IRF1, TYMP) characteristic genes. Figure 1 The F in the selected characteristic genes were marked in the volcano plot of GSE248830 (F). Figure 1 (G in the middle).

[0149] Example 2: Consistent Cluster Analysis Based on Characteristic Genes of Lung Adenocarcinoma Brain Metastases

[0150] Consistent cluster analysis was performed on the transcriptome data GSE248830 using five characteristic genes of lung adenocarcinoma brain metastasis. The results are as follows: Figure 2 As shown.

[0151] The results show that when k=2, the clustering stability is higher. Figure 2 A- Figure 2 (C) The Kaplan-Meier survival curves show that cluster B has a better overall survival (OS) and progression-free survival (PFS) compared to cluster A. Figure 2 D- Figure 2As shown in E in the figure. Further investigation into the expression of brain metastasis characteristic genes in clusters A and B revealed that BNIP3 was expressed at a higher level in cluster A, while TYMP, CD74, IRF1, and LYZ were expressed at a higher level in cluster B. Figure 2 (F in the middle).

[0152] KEGG enrichment analysis was used to identify the enrichment pathways for each cluster. Figure 2 The study aimed to explore the functional pathways of clusters A and B. Results showed that antigen processing and presentation, B cell receptor signaling pathways, chemokine signaling pathways, natural killer cell-mediated cytotoxicity, and T cell receptor signaling pathways were enriched in cluster B, while ascorbic acid and aldose metabolism, cell cycle, cytochrome P450 exogenous substance metabolism, pentose-glucuronic acid interconversion, porphyrin and chlorophyll metabolism, retinol metabolism, and steroid hormone biosynthesis were enriched in cluster A.

[0153] Example 3: Tumor immune microenvironment and tumor mutation burden analysis of clusters A and B

[0154] The role of the tumor immune microenvironment in brain metastasis of lung adenocarcinoma was further investigated by calculating immune, matrix, and ESTIMATE scores through ESTIMATE analysis. Figure 3 As shown.

[0155] The results showed that the scores of cluster B were significantly higher than those of cluster A. Figure 3 The presence of A in the data indicates that brain metastasis in lung adenocarcinoma is associated with reduced immune cell infiltration. Furthermore, CIBERSORT (a method for estimating relative proportions of cell subsets based on RNA transcripts) and ssGSEA (single-sample gene set enrichment analysis) were used to assess the differences in immune cell subset infiltration between the two clusters. Significant differences were found in the infiltration levels of 13 immune cell subsets between the two clusters, and the scores of immune-related pathways in cluster B were significantly higher than those in cluster A. Figure 3 B- Figure 3 (C in the middle).

[0156] Subsequently, the tumor mutational burden of clusters A and B was investigated. The results showed that the tumor mutational burden level in cluster A was significantly higher than that in cluster B. Figure 3 (D in the diagram); The waterfall plot shows the mutation rates of the two clusters, with cluster A having a higher mutation rate ( Figure 3 E in the context. Therefore, patients are divided into four distinct categories based on their tumor mutation burden and clustering (E in the context). Figure 3 The results showed that individuals with high tumor mutational burden in cluster B were most likely to achieve good clinical outcomes, while individuals with low tumor mutational burden in cluster A had the worst overall survival.

[0157] Tumor immune microenvironment and tumor mutation burden analysis affect tumor sensitivity to drug therapy; therefore, the sensitivity of clusters A and B to anticancer drugs was further evaluated. Results showed that 5-Fluorouracil, Afatinib, Erlotinib, Gefitinib, and Savolitinib showed higher sensitivity in cluster A, while patients in cluster B received greater clinical benefit from axitinib treatment. Figure 3 (G in the middle).

[0158] Example 4: Cell communication analysis reveals the mechanism by which BNIP3, CD74, and LYZ regulate lung adenocarcinoma brain metastasis through CellChat in epithelial cells and other cells.

[0159] To explore whether the high expression of BNIP3 and the low expression of CD74, LYZ, IRF1, and TYMP in epithelial cells are consistent with altered cell communication in brain metastases, the CellChat database was used to analyze differences in cell communication among various cell types in brain metastases, such as... Figure 4 As shown. Based on whether they express BNIP3, epithelial cells are divided into BNIP3+ epithelial cells and BNIP3- epithelial cells. Studies have found that, compared to BNIP3- epithelial cells, BNIP3+ epithelial cells exhibit stronger interactions with endothelial cells, mast cells, myeloid cells, NK cells, T cells, and B cells. Figure 4 (A) Furthermore, the interaction from fibroblasts to BNIP3+ epithelial cells was significantly stronger than that to BNIP3- epithelial cells. However, for CD74+ epithelial cells, the interaction from fibroblasts to endothelial cells, mast cells, myeloid cells, NK cells, T cells, and B cells was weaker than that to CD74- epithelial cells, while the interaction to fibroblasts was stronger than that to CD74- epithelial cells. Figure 4 (B in the original text). Furthermore, the signal intensity emitted by fibroblasts, endothelial cells, and oligodendrocytes to CD74+ epithelial cells was significantly stronger than that to CD74- epithelial cells. Similarly, for LYZ+ epithelial cells, the interaction between their signals and those emitted to endothelial cells, mast cells, myeloid cells, NK cells, T cells, and B cells was weaker than that to LYZ- epithelial cells, while the signal intensity emitted by fibroblasts, endothelial cells, and oligodendrocytes to LYZ+ epithelial cells was weaker than that to LYZ- epithelial cells. Figure 4 (C) Similarly, analysis of the interactions between IRF1+ / -, TYMP+ / - epithelial cells and other cells showed no significant differences. Likewise, as... Figure 5 As shown, analysis of the interactions between IRF1 + / - and TYMP+ / - epithelial cells and other cells revealed no significant differences.

[0160] In summary, combining the expression of BNIP3, CD74, and LYZ, epithelial cells were divided into BNIP3+CD74-LYZ- epithelial cells and other epithelial cells, with more significant differences. Figure 4 As shown in D, the interactions from BNIP3+CD74-LYZ- epithelial cells to endothelial cells, mast cells, myeloid cells, NK cells, T cells, and B cells are significantly stronger than those from other epithelial cells. Furthermore, the interactions from endothelial cells and oligodendrocytes to BNIP3+CD74-LYZ- epithelial cells are weaker than those to other epithelial cells.

[0161] Further in-depth analysis of intercellular receptor signaling pathways is needed. For example... Figure 4 As shown in E, BNIP3+CD74-LYZ- epithelial cells primarily communicate with endothelial cells via the MDK signaling pathway, and with other cell types primarily via the MIF signaling pathway [mainly via the MIF-(CD74+CXCR4) pathway to connect with NK cells, T cells, and B cells, and mainly via the MIF-(CD74+CD44) pathway to connect with mast cells and myeloid cells]. Therefore, further exploration of the role of the MIF signaling pathway in intercellular communication yielded the following results: Figure 4 As shown in F, the MIF signaling emitted by BNIP3+CD74-LYZ- epithelial cells was significantly stronger than that of other epithelial cells. Due to the significant differences in MIF signaling pathway communication among epithelial cells, further investigation was conducted into the ligand-receptor pairs in the MIF pathway, which plays a major role in communication between epithelial cells and other cells. Figure 4 The results showed that BNIP3+CD74-LYZ- epithelial cells were stronger than other epithelial cells in both MIF-(CD74+CXCR4) and MIF-(CD74+CD44) ligand-receptor pairs, and that BNIP3+CD74-LYZ- epithelial cells showed increased expression of ligands and receptors in signaling pathways during interactions with other cells.

[0162] Example 5: Classification of T cells, B cells, and myeloid cell subsets and analysis of intercellular communication with BNIP3+CD74-LYZ- epithelial cells

[0163] To further elucidate the cell subsets involved in the regulation of lung adenocarcinoma brain metastasis by BNIP3+CD74-LYZ- epithelial cells, T cells, B cells, and myeloid cells were subgrouped, such as... Figures 6-8 As shown, T cells were clustered and annotated into four cell populations: Treg (regulatory T cells), NKT (natural killer T cells), Native CD4+ T cells, and CD8+ T cells, as shown. Figure 6As shown in A. Further research revealed that CD8+ T cells, as the main T cells, receive communication signals from BNIP3+CD74-LYZ- epithelial cells ( Figure 6 In B). Communication between BNIP3+CD74-LYZ- epithelial cells and T cell subsets mainly relies on the MIF−(CD74+CXCR4) ligand-receptor pair, and its signal is stronger with the ligand-receptor pair of CD8+ T cells ( Figure 6 (C in the text). Subsequently, heatmaps were used to identify the signals that contributed most to the output or input signals of BNIP3+CD74-LYZ- epithelial cells and T cell subsets. Six signaling pathways were highly active and played major roles: MIF, MHC-I, APP, MK, CDH1, and LAMININ pathways (…). Figure 6 (D in the text). Consistent with previous results, the MIF signaling pathway is most strongly involved in communication between BNIP3+CD74-LYZ- epithelial cells and CD8+ T cells, with the MIF-(CD74+CXCR4) and MIF-(CD74+CD44) receptor pairs playing a major role. Figure 6 E- Figure 6 (G in the middle).

[0164] B cells are further subdivided into three subsets: follicular B cells, plasma cells, and MALT B cells. Figure 7 A in the text). BNIP3+CD74-LYZ- epithelial cells mainly communicate with follicular B cells ( Figure 7 In B), communication between epithelial cells and follicular B cells and plasma cells is mainly through MIF−(CD74+CXCR4), while communication with MALT B cells is mainly through APP-CD7 (…). Figure 7 In addition, the six signaling pathways of MIF, CD99, APP, MK, CDH1, and LAMININ contribute the most to inter-cell communication. Figure 7 D). The MIF signaling pathway is strongest in cell communication from BNIP3+CD47−LYZ− epithelial cells and plasma cells to follicular B cells. Figure 7 E and Figure 7 The MIF pathway plays a key role in cell communication between BNIP3+CD74-LYZ-epithelial cells and follicular B cells, while the role of the MIF pathway in BNIP3+CD74-LYZ-epithelial cells and MALT B cells is mediated by MIF-(CD74+CXCR4). Figure 7 (G in the middle).

[0165] Finally, myeloid cells were re-clustered into 5 subgroups: monocytes, monocyte-macrophages, microglia, alveolar macrophages, and indeterminate ( Figure 8 (A in the text). Cell communication results revealed that the interaction from BNIP3+CD47−LYZ− epithelial cells to microglia was slightly stronger than that to other myeloid cell subsets (A in the text). Figure 8 In the signaling pathway from BNIP3+CD47-LYZ- epithelial cells to myeloid cell subsets, the three ligand-receptor pairs MIF-(CD74+CXCR4), MIF-(CD74+CD44), and APP-CD74 have relatively strong effects. Figure 8 (C) in the text. Figure 8 As shown in D, the seven signaling pathways of MIF, CD99, APP, MK, GRN, LAMININ, and ANNEXIN are more active. The MIF signaling pathway mainly plays a role in the interaction between BNIP3+CD47-LYZ- epithelial cells and alveolar macrophages and microglia. Figure 8 E- Figure 8 Finally, further exploration of the receptor-ligand pairs in the MIF pathway revealed that MIF-(CD74+CXCR4) also primarily plays a role in the interaction between BNIP3+CD47-LYZ- epithelial cells and alveolar macrophages and microglia, while the intensity of MIF-(CD74+CD44) interaction was relatively consistent across groups. Figure 8 (G in the middle).

[0166] Example 6: Epithelial Cell Reclustering and Functional Enrichment Analysis

[0167] Epithelial cells were extracted from brain metastases and then subjected to cluster analysis, such as... Figure 9 As shown. Figure 9 As shown in A, the epithelial cells are further subdivided into 6 clusters (C0-C5); as Figure 9 As shown in Figure B, the heatmap displays the marker molecules specifically expressed in each cell cluster. Subsequently, the expression of BNIP3, CD74, and LYZ in each cell cluster was investigated. BNIP3 was mainly expressed in C1 and C4, CD74 was mainly expressed in C0, C2, C3, and C5, while LYZ was expressed in small amounts in C0, C2, and C5. Figure 9(C in the text). Based on the above results, BNIP3+CD47-LYZ- epithelial cells were defined as positive cells, and other epithelial cells were defined as negative cells. Interestingly, the positive epithelial cells were mainly C1 and C4 clusters. Furthermore, two subsets of epithelial cells were extracted (positive and negative), and the proportions of the six epithelial cell subclusters (C0-C5) were calculated separately. Similarly, in the positive epithelial cells, the proportions of C1 and C4 clusters increased significantly, while the proportions of C0, C2, and C3 clusters decreased significantly, and the proportion of C5 clusters remained relatively unchanged. Figure 9 (D in the text). Finally, we focused on C1 and C4 cluster epithelial cells, as they may be closely related to brain metastasis of lung adenocarcinoma. Differential analysis was performed on C1, C4, and other cluster epithelial cells, and enrichment analysis was performed on genes highly expressed in C1 and C4 clusters. In KEGG enrichment, pathways related to distant tumor metastasis (PI3K-Akt signaling pathway, focal adhesion, autophagy, VEGF signaling pathway, HIF-1 signaling pathway, Hippo signaling pathway, cytoskeleton in muscle cells, etc.) were activated. Similarly, in GO enrichment, MF entries (membrane receptor protein tyrosine kinase activator activity), CC entries (contractile fibers, actin filaments), and BP entries (response to hypoxia, regulation of cell-matrix adhesion, cellular response to hypoxia) related to distant tumor metastasis were activated. Figure 9 E- Figure 9 (F in the middle).

[0168] Example 7: Tumor progression-related gene set scoring in epithelial cells

[0169] By integrating expression information from multiple genes to assess biological functions or pathway activity related to tumor progression, such as... Figures 10-11 As shown. First, tumor metastasis and proliferation were scored for different epithelial cell subsets. It was found that the C1 cluster of epithelial cells had significantly stronger metastatic ability than other clusters, while the C4 cluster of epithelial cells had the strongest proliferative ability. Figure 10 A- Figure 10 (B) This is consistent with the finding in Example 6 that the epithelial cells of clusters C1 and C4 are epithelial cells associated with brain metastasis. Figure 10 As shown in C, BNIP3 expression was positively correlated with proliferation and migration scores, while CD74 and LYZ were negatively correlated.

[0170] Subsequently, heatmaps were used to display the activity scores of some common pathways, focusing on the biological pathways activated in C1 and C4 cluster epithelial cells. For example... Figure 10As shown in D, the hypoxia activity of the C1 and C4 clusters was significantly higher than that of other clusters. The C4 cluster showed higher activity of the notch signaling pathway, the epithelial-mesenchymal transition pathway, and the notch signaling pathway, while the C1 cluster showed higher activity of the notch signaling pathway, the notch signaling pathway, and the notch signaling pathway.

[0171] Furthermore, Kaplan-Meier survival analysis based on the TCGA database, such as Figure 11 As shown, the results indicated that high BNIP3 expression was associated with poorer overall survival (OS) in patients with lung adenocarcinoma, and may be associated with poorer progression-free survival (PFS) in patients with lung adenocarcinoma, but the difference was not statistically significant. Figure 11 As shown in A; low CD74 expression was significantly associated with poorer overall survival and progression-free survival in patients with lung adenocarcinoma, such as... Figure 11 As shown in B; low LYZ expression may be associated with poorer overall survival and progression-free survival in patients with lung adenocarcinoma, such as... Figure 11 As shown in C.

[0172] Example 8: Pseudo-temporal analysis of brain metastasis-related epithelial cells

[0173] To elucidate the role of epithelial cells in brain metastasis of lung adenocarcinoma, a branching trajectory was constructed using pseudo-temporal analysis to reveal the continuous patterns of cell state evolution, such as... Figure 12 As shown. Figure 12 As shown in A, the trajectory diagram is drawn according to the pseudo-time value, cell state, cell type, and Seurat group. When states 1, 2, and 5 differentiate and develop to node 1, two different branches appear (state 3 and state 4), in which state 4 is mainly enriched in C1 and C4 cluster epithelial cells.

[0174] Subsequently, cell density maps along the time axis suggested that both C1 clusters and BNIP3+CD74−LYZ− epithelial cells were in the late stage of epithelial cell development. Figure 12 (B in the original text). Furthermore, studies on the expression changes of brain metastasis-related marker genes BNIP3, CD74, and LYZ with different cellular states revealed that BNIP3 expression increases with cell development, CD74 expression decreases with cell development, and LYZ expression shows no significant change. Figure 12 (C in the original text). To identify genes that change with cell progression, a branch-dependent gene heatmap was created. The genes in the heatmap were categorized into four clusters. Genes in cluster 1 showed increased expression as prebranching progressed towards cell fate 2, while genes in cluster 3 showed increased expression as prebranching progressed towards cell fate 1. This indicates that genes in cluster 1 regulate the differentiation of epithelial cells from prebranching to cell fate 2, while genes in cluster 3 regulate differentiation towards cell fate 1. Figure 12(D in the original text). Furthermore, cell fate 2 was primarily composed of C1 and C4 cluster epithelial cells, which are closely associated with brain metastasis. Finally, genes from clusters 1 and 3 were extracted and subjected to GO and KEGG enrichment analyses, respectively. The enrichment results are shown below. Figure 12 E and Figure 12 As shown in F in the diagram. In the GO enrichment, genes regulating differentiation towards cell fate 2 are enriched in responses to hypoxia, reduced oxygen levels, oxygen concentrations, focal adhesion, cell-matrix junctions, and cadherin binding, which are closely related to distant tumor metastasis. Genes regulating differentiation towards cell fate 1 are enriched in regulating T cell activation, immune response activation signaling pathways, leukocyte cell-cell adhesion, MHC protein binding, antigen binding, and MHC protein complex binding, which are closely related to anti-tumor immunity. In the KEGG enrichment, genes regulating differentiation towards cell fate 2 are enriched in the HIF-1 signaling pathway, PI3K-Akt signaling pathway, and autophagy, which are closely related to distant tumor metastasis. Genes regulating differentiation towards cell fate 1 are enriched in antigen processing and presentation, natural killer cell-mediated cytotoxicity, and T cell receptor signaling pathways, which are closely related to anti-tumor immunity.

[0175] Example 9: Clinical specimen verification of the correlation between BNIP3, CD74, LYZ and brain metastasis and prognosis in patients with lung adenocarcinoma

[0176] 1. Immunohistochemistry (IHC) was used to detect the expression of BNIP3, CD74, and LYZ in brain metastasis tissues from 54 patients with lung adenocarcinoma brain metastases. Simultaneously, primary lung adenocarcinoma tumor tissues were obtained from 8 of these patients. The expression of BNIP3, CD74, and LYZ in the primary lung adenocarcinoma tumor tissues and brain metastasis tissues of paired patients with lung adenocarcinoma brain metastases was analyzed. Figure 13 As shown in A in the diagram.

[0177] The results showed that BNIP3 and LYZ were expressed in the cytoplasm, and CD74 was expressed in both the cytoplasm and cell membrane. Furthermore, the expression of BNIP3 in brain metastases was higher than that in the corresponding primary lung adenocarcinoma tumor tissue, while the expression of CD74 in brain metastases was lower than that in the corresponding primary lung adenocarcinoma tumor tissue. In contrast, the expression of LYZ was low in both the primary lung adenocarcinoma tumor tissue and brain metastases.

[0178] 2. Brain metastasis tissues from patients with lung adenocarcinoma and brain metastases were used as the experimental group. Primary tumor tissues and paired adjacent normal tissues from 19 patients with lung adenocarcinoma without brain metastases, 18 schwannomas, and 19 gliomas were used as the four control groups. Specifically, a BNIP3 immunohistochemical score ≤1.7 was defined as negative, 1.7 < score ≤4 as weakly positive, 4 < score ≤8 as positive, and a score >8 as strongly positive; a CD74 immunohistochemical score ≤1 was defined as negative, 1 < score ≤2.3 as weakly positive, 2.3 < score ≤6.7 as positive, and a score >6.7 as strongly positive; and a LYZ immunohistochemical score =0 was defined as negative, 0 < score ≤0.7 as weakly positive, 0.7 < score ≤2.3 as positive, and a score >2.3 as strongly positive.

[0179] The comparison results are as follows Figure 14 As shown, BNIP3 expression was significantly higher in the experimental group than in the control group, indicating high expression specifically in the brain metastases of patients with lung adenocarcinoma, while CD74 expression was significantly lower in the experimental group than in the control group. Furthermore, as... Figure 13 As shown in B, BNIP3 and LYZ are hardly expressed in primary lung adenocarcinoma tissues without metastasis, adjacent normal tissues, schwannomas, and gliomas, while CD74 is expressed at a low level.

[0180] 3. Multiple immunofluorescence staining was performed on tissues (brain metastasis tissue, primary lung adenocarcinoma tissue, primary lung adenocarcinoma tissue without brain metastasis, and adjacent normal tissue) from 12 patients with lung adenocarcinoma with brain metastasis and 12 patients with lung adenocarcinoma without brain metastasis. Tumor epithelial cells were labeled with EPCAM (epithelial cell adhesion molecule).

[0181] Multiple immunofluorescence staining of brain metastatic tissue, primary lung adenocarcinoma tissue, primary lung adenocarcinoma tissue without metastasis, and adjacent normal tissue, such as... Figure 13 As shown in C, the proportion of brain metastasis-associated epithelial cells co-localized by EPCAM+BNIP3+CD74-LYZ staining is higher in brain metastasis tissue than in primary lung adenocarcinoma tissue, and the proportion of these cells is higher in primary lung adenocarcinoma tissue with metastasis than in primary lung adenocarcinoma tissue without metastasis. However, they are almost absent in adjacent normal tissue.

[0182] 4. Further analysis was conducted on the relationship between the expression of BNIP3, CD74, and LYZ and the clinicopathological characteristics of patients with brain metastases from lung adenocarcinoma, as shown in Table 1. BNIP3 was defined as low expression with an immunohistochemical score ≤6.7 and high expression with a score >6.7; CD74 was defined as low expression with an immunohistochemical score ≤1 and high expression with a score >1; and LYZ was defined as low expression with an immunohistochemical score ≤0.5 and high expression with a score >0.5. The results showed that the expression of BNIP3 and LYZ was correlated with N stage, indicating that BNIP3 and LYZ are potential biomarkers for brain metastases in patients with lung adenocarcinoma and can help in the timely early warning of metastatic behavior.

[0183] Table 1. Tissue expression of BNIP3, CD74, and LYZ and clinical characteristics of patients with brain metastases from lung adenocarcinoma.

[0184]

[0185] Further analysis of the relationship between the expression of BNIP3, CD74, and LYZ and tumor markers is shown in Table 2. The results showed that patients with high CD74 expression had lower levels of CYFRA21-1 in their blood than those with low CD74 expression, and patients with high LYZ expression had lower levels of CA125 in their blood than those with low LYZ expression, and these differences were statistically significant.

[0186] Table 2. Correlation between tissue expression of BNIP3, CD74, and LYZ and tumor marker levels.

[0187]

[0188] Note: CEA, carcinoembryonic antigen; CA125, carbohydrate antigen 125; CA199, carbohydrate antigen 199; CYFRA21-1, cytokeratin 19 fragment; SF, serum ferritin.

[0189] Finally, to investigate the prognostic significance of BNIP3, CD74, and LYZ in patients with brain metastases from lung adenocarcinoma, the relationship between their expression and overall survival and progression-free survival after brain metastases in lung adenocarcinoma patients was analyzed. Figure 13As shown in D in the figure. Based on the quantitative results of BNIP3 immunohistochemistry, lung adenocarcinoma patients with brain metastases were divided into high-expression and low-expression groups. Kaplan-Meier survival curves showed that the overall survival after brain metastases was shorter in the high-expression BNIP3 group than in the low-expression BNIP3 group (high-expression BNIP3: 24.6 months, low-expression BNIP3: 37.1 months, P = 0.038). Similarly, the progression-free survival after brain metastases was shorter in the high-expression BNIP3 group than in the low-expression BNIP3 group (high-expression BNIP3: 9.5 months, low-expression BNIP3: 16.5 months, P = 0.083). For CD74, the overall survival after brain metastases was longer in the high-expression CD74 group than in the low-expression CD74 group (high-expression CD74: 37.1 months, low-expression CD74: 22.1 months, P = For LYZ, overall survival after brain metastasis was longer in the high-expression group than in the low-expression group (18.1 months vs. 9.4 months, P = 0.036); for LYZ, overall survival after brain metastasis was longer in the high-expression group than in the low-expression group (32.8 months vs. 22.1 months, P = 0.24), and progression-free survival after brain metastasis was also longer in the high-expression group than in the low-expression group (20.6 months vs. 9.5 months, P = 0.0028).

[0190] A combined analysis of the correlation between BNIP3, CD74, and LYZ expression and the prognosis of patients with brain metastases from lung adenocarcinoma revealed that patients with high BNIP3 expression, low CD74 expression, and low LYZ expression had significantly shorter overall survival and progression-free survival after brain metastases compared to other patients. The results showed that overall survival was 16.0 months in the high BNIP3, low CD74, and low LYZ expression group, compared to 35.2 months in the other patient group (P < 0.0001); progression-free survival was 6.8 months in the high BNIP3, low CD74, and low LYZ expression group, compared to 16.5 months in the other patient group (P < 0.0001).

[0191] Example 10:

[0192] To further clarify the diagnostic predictive efficacy of BNIP3, CD74, and LYZ for patients with brain metastases from lung adenocarcinoma, and to test their ability to differentiate between primary lung adenocarcinoma lesions with and without brain metastases, this application plotted ROC curves and calculated the AUC (area under the curve). Figure 15As shown, the AUC of the combined BNIP3, CD74, and LYZ model was 0.868 (accuracy: 93.1%; sensitivity: 80.0%; specificity: 100.0%), which is superior to the AUC of BNIP3 (0.811, accuracy: 79.3%; sensitivity: 90.0%; specificity: 73.7%), CD74 (0.608, accuracy: 65.5%; sensitivity: 90.0%; specificity: 52.6%), and LYZ (0.621, accuracy: 58.6%; sensitivity: 100%; specificity: 36.8%).

[0193] Based on Examples 1-10, this invention identified brain metastasis markers BNIP3, CD74, and LYZ through bioinformatics analysis. Through single-cell analysis of cell interactions in the tumor microenvironment of lung adenocarcinoma brain metastasis, a group of epithelial cells closely related to brain metastasis was also discovered: BNIP3+CD74-LYZ-epithelial cells.

[0194] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. Use of a reagent for detecting a lung adenocarcinoma brain metastasis diagnostic marker combination of BNIP3, CD74 and LYZ in the preparation of a reagent for judging the prognosis of lung adenocarcinoma brain metastasis, wherein the reagent is used for judging the prognosis of lung adenocarcinoma brain metastasis by detecting the expression levels of BNIP3, CD74 and LYZ in brain metastasis tissues of lung adenocarcinoma brain metastasis patients.

2. Use according to claim 1, characterized in that: The reagent comprises a reagent for detecting the protein expression level and / or mRNA expression level of BNIP3, CD74 and LYZ.

3. Use according to claim 2, wherein: The reagent for detecting the protein expression level of BNIP3, CD74 and LYZ comprises an anti-BNIP3 antibody, an anti-CD74 antibody and an anti-LYZ antibody for immunohistochemistry, immunofluorescence or Western blotting.

4. Use according to claim 3, wherein: The reagent for detecting the mRNA expression level of BNIP3, CD74 and LYZ comprises a reagent for qRT-PCR, RNA sequencing or fluorescence in situ hybridization.

5. Use of an antibody for detecting BNIP3, CD74 and LYZ in the preparation of a kit for diagnosing lung adenocarcinoma brain metastasis.

6. Use according to claim 5, wherein: The antibody for detecting BNIP3, CD74 and LYZ comprises an anti-BNIP3 antibody, an anti-CD74 antibody and an anti-LYZ antibody for immunohistochemistry or immunofluorescence.

Images