Prognostic model related to triple-negative breast cancer pyroptosis and construction method and application thereof

Abstract

The invention provides a triple-negative breast cancer pyroptosis related prognosis model and a construction method and application thereof, and the construction method of the model comprises the following steps: (1) obtaining transcriptome data and prognosis data of triple-negative breast cancer patients and normal people from a database, and analyzing to obtain triple-negative breast cancer differential expression genes; and (2) screening from the database to obtain an intersection of the pyroptosis related genes and the differentially expressed genes, based on the genes in the intersection, screening through Lasso Cox regression analysis to obtain genes related to triple negative breast cancer prognosis, and using the genes to construct a risk scoring model. The model can effectively predict the prognosis of the triple negative breast cancer patient through risk scoring.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, specifically to a pyroptosis-related prognostic model for triple-negative breast cancer, its construction method, and its application. Background Technology

[0002] Breast cancer has become the most common cancer worldwide, threatening the health of women globally. Among all molecular subtypes of breast cancer, triple-negative breast cancer is characterized by its high invasiveness, high mortality rate, high recurrence and metastasis rate, and poor prognosis, making it a persistent challenge in the diagnosis and treatment of breast cancer. Although progress has been made in the treatment of triple-negative breast cancer in recent years, a significant number of patients still do not benefit from new adjuvant chemotherapy or immune checkpoint suppression therapy. Therefore, there is an urgent need to identify new and appropriate biomarkers to construct models for predicting the prognosis of triple-negative breast cancer, thereby enabling the selection of the optimal treatment regimen to improve patient outcomes.

[0003] The tumor immune microenvironment plays an irreplaceable role in the proliferation, invasion, and metastasis of tumor cells and is a key factor determining patient treatment response and prognosis. In recent years, pyroptosis has received widespread attention due to its close relationship with tumor-infiltrating immune cells. Pyroptosis is a programmed cell death mediated by inflammasomes, characterized by the release of inflammatory cytokines. This process plays a crucial role in the immune response; even a small number of cells undergoing pyroptosis are sufficient to trigger an inflammatory response, regulate tumor-infiltrating immune cells, and activate a robust T-cell anti-tumor immune response, synergizing with immune checkpoint inhibitors to enhance anti-tumor efficacy. However, considering its inflammatory nature, aberrant pyroptosis may also be related to the formation of a tumor-supporting microenvironment. Tumor cells may utilize pyroptosis mechanisms to evade the immune system's attack, suppress the immune response, and thus promote their own survival and proliferation. Therefore, the heterogeneity of tumor-infiltrating immune cells caused by pyroptosis poses a significant challenge to tumor treatment and prognosis. However, the complex roles and prognostic significance of pyroptosis-related genes in triple-negative breast cancer remain unclear, and there is no relevant research on the potential of pyroptosis in anti-tumor immunity in triple-negative breast cancer.

[0004] Therefore, exploring the relationship between pyroptosis-related genes and triple-negative breast cancer and constructing predictive models or tools for triple-negative breast cancer based on pyroptosis-related genes is of great significance for better predicting the progression, prognosis and treatment effects of triple-negative breast cancer. Summary of the Invention

[0005] To address the problems existing in the background art, the present invention provides a pyroptosis-related prognostic model for triple-negative breast cancer, its construction method, and its application. This model can effectively predict the prognosis of triple-negative breast cancer patients through risk scoring.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides a method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer, comprising the following steps: (1) Transcriptome data and prognostic data of triple-negative breast cancer patients and normal individuals were obtained from the database, and differentially expressed genes of triple-negative breast cancer were obtained by comparing transcriptome data. (2) The intersection of pyroptosis-related genes and the differentially expressed genes was obtained from the database. Based on the genes in the intersection, genes related to the prognosis of triple-negative breast cancer were screened by Lasso Cox regression analysis and used to construct a risk scoring model. The risk scoring model is expressed as follows: Risk Score=β1×Exp(mRNA1)+β2×Exp(mRNA2)+...+β n ×Exp(mRNAn) Where β1, β2, ..., β n is the coefficient of the gene, and mRNA1, mRNA2...mRNAn are the expression levels of the gene; According to the above scheme, the genes related to the prognosis of triple-negative breast cancer obtained in step (2) are: PINK1, GZMB, PFKFB3, RSPO3, TREM1 and VEGFA.

[0007] According to the above scheme, the risk scoring model in step (2) is as follows: RiskScore = (-0.15836×GZMB expression level) + (0.32236×PINK1 expression level) + (0.11124×PFKFB3 expression level) + (-0.04646×RSPO3 expression level) + (0.04374×TREM1 expression level) + (0.03647×VEGFA expression level) (C Index: 0.722).

[0008] According to the above scheme, in step (2), the prognostic model is evaluated based on the training set data and the test set data.

[0009] According to the above plan, the evaluation method is as follows: Calculate the risk score for each subject in the training dataset based on the risk scoring model; Scoring was analyzed using Kaplan-Meier curves from the training and test sets, as well as time-dependent receiver operating characteristic curves. The predictive efficacy of the model was evaluated using Kaplan-Meier curves from the training and test sets, as well as time-dependent receiver operating characteristic (ROC) curves.

[0010] According to the above plan, based on the constructed risk scoring model, patients are divided into high-risk and low-risk groups according to the median value.

[0011] Secondly, the present invention provides a pyroptosis-related prognostic model for triple-negative breast cancer constructed by the above method.

[0012] According to the above scheme, the area under the ROC curve for the five-year survival of patients with triple-negative breast cancer using the prognostic model reached 0.83.

[0013] Thirdly, this invention provides the application of the above-mentioned triple-negative breast cancer pyroptosis-related prognostic model in assessing the prognosis of triple-negative breast cancer patients.

[0014] According to the above scheme, the expression level of genes related to prognosis is the expression level of genes related to prognosis in tumor samples from triple-negative breast cancer patients.

[0015] The beneficial effects of this invention are: This invention explores the relationship between pyroptosis-related genes and the prognosis of triple-negative breast cancer (TNC). Novel prognostic biomarkers associated with pyroptosis were screened and validated from massive datasets, and a pyroptosis-related prognostic model for TNC was successfully constructed. This model more accurately reflects the disease progression and prognosis of TNC patients from a pyroptosis perspective. ROC curve analysis showed an AUC value of 0.83 for five-year survival, and immunohistochemical validation confirmed a high degree of consistency with the model score, demonstrating high diagnostic efficacy. This prognostic model is significant for optimizing clinical decision-making, enabling personalized medical strategies, and improving patient survival and quality of life, and has the potential to become a novel auxiliary reference standard for the management of TNC. Attached Figure Description

[0016] Figure 1 Figure 1 is a schematic diagram of the screening of pyroptosis-related genes in this invention (A: Venn diagram of the intersection of differentially expressed genes and pyroptosis-related genes; B: Volcano diagram of differentially expressed pyroptosis-related genes; C: Heatmap of differentially expressed pyroptosis-related genes; D: GO enrichment analysis of differentially expressed pyroptosis-related genes; E: KEGG enrichment analysis of differentially expressed pyroptosis-related genes). Figure 2 The graphs show the KM curves and time-dependent ROC curves, where A represents the KM curves of the high and low risk groups in the training set, B represents the ROC curves of the high and low risk groups in the training set, C represents the KM curves of the high and low risk groups in the validation set, and D represents the ROC curves of the high and low risk groups in the validation set.

[0017] Figure 3 The correlation of model genes and their expression in single cells are shown in Figure 1, where A represents the correlation of model genes and BG represents the expression of model genes in single cells. Figure 4The spatial expression of the model genes of this invention is shown in Figure AB: dimensionality-reduced Umap map of spatial transcriptomics data and spatial distribution map of each cluster; CD: annotated Umap map and spatial distribution of each cell type; EG: spatial distribution of the 6 model genes. Figure 5 The immunohistochemical results are for pyroptosis-related prognostic genes of this invention. Detailed Implementation

[0018] The principles and features of the present invention are described below with reference to the accompanying drawings and specific embodiments. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0019] Example 1 1. Acquisition of transcriptome data from triple-negative breast cancer patients Transcriptomic and clinical information of 97 triple-negative breast cancer patients were obtained from the TCGA (Cancer Genome Atlas) database. Transcriptomic and clinical information of 135 triple-negative breast cancer patients were obtained from the GEO (Gene Expression Omnibus) database. After batch effect removal and data standardization, the two cohorts were merged into a single cohort of 232 patients. This cohort included 223 tumor samples and 9 adjacent normal tissue samples as a training cohort; additionally, transcriptomic and clinical information of 154 triple-negative breast cancer patients were obtained from the GEO database as a validation cohort.

[0020] 2. Obtain differentially expressed genes in triple-negative breast cancer The acquired transcriptome data were analyzed using the limma package in R. Differentially expressed genes in triple-negative breast cancer were identified through differential transcriptome analysis between tumor and normal samples. The selection criteria were a corrected p-value < 0.05 and log2|FC| > 1.

[0021] 3. Further screen differentially expressed genes related to pyroptosis. By intersecting the pyroptosis-related genes collected from GeneCards, MSigDB databases, and literature with the above-mentioned differentially expressed genes, 61 differentially expressed genes related to pyroptosis in triple-negative breast cancer were obtained; GO enrichment and KEGG enrichment analyses were performed on these 61 genes. Figure 1 This diagram illustrates the screening of pyroptosis-related genes. A is the Venn diagram of the intersection of differentially expressed genes and pyroptosis-related genes; B is the volcano diagram of differentially expressed pyroptosis-related genes; C is the heatmap of differentially expressed pyroptosis-related genes; D is the GO enrichment analysis of differentially expressed pyroptosis-related genes; and E is the KEGG enrichment analysis of differentially expressed pyroptosis-related genes.

[0022] 4. Model building and validation In the training cohort, Lasso Cox regression was used to model the transcriptome data of triple-negative breast cancer. The model formula was: Risk Score = β1 × Exp(mRNA1) + β2 × Exp(mRNA2) + ... + β n ×Exp(mRNAn), Where β1, β2...β n is the coefficient of the gene, and mRNA1, mRNA2...mRNAn are the expression levels of the gene.

[0023] A prognostic risk model consisting of six genes was obtained: PINK1, GZMB, PFKFB3, RSPO3, TREM1, and VEGFA. The final model is as follows: RiskScore = (-0.158360246047182 × GZMB expression level) + (0.32235503219597 × PINK1 expression level) + (0.111243442274487 × PFKFB3 expression level) + (-0.0464626296260152 × RSPO3 expression level) + (0.0437384204335007 × TREM1 expression level) + (0.0364743584502037 × VEGFA expression level) The risk score for each patient was calculated using the above formula, and patients were divided into high-risk and low-risk groups based on the median of all scores. Model efficiency was tested in a validation cohort using packages and related techniques such as "glmnet", "survival", "ggplot2", "survminer", and "timeROC". Survival analysis between the high-risk and low-risk groups on both the training and validation sets showed significant differences. KM curves and time-dependent ROC curves were plotted, and the results are as follows: Figure 2 As shown, A represents the KM curves for the high- and low-risk groups in the training set, B represents the ROC curves for the high- and low-risk groups in the training set, C represents the KM curves for the high- and low-risk groups in the validation set, and D represents the ROC curves for the high- and low-risk groups in the validation set. The area under the ROC curve for the model's five-year overall survival in the validation set reaches 0.83, indicating that the model has good predictive performance.

[0024] Single-cell sequencing data from 10 cases of triple-negative breast cancer and 4 cases of normal breast tissue were obtained from GEO. Data preprocessing was performed, and the correlation and expression of six model genes in single cells were analyzed. The correlation and expression of each gene in single cells are shown below. Figure 3As shown, A represents the correlation of model genes, and BG represents the expression of model genes in single cells. The results, demonstrated using biological methods, show that the six genes in the prognostic risk model can be integrated and effectively predicted because of their correlation and the differences in their expression across different cells. These genes collectively influence the tumor immune microenvironment in triple-negative breast cancer patients, thereby affecting their prognosis through pyroptosis.

[0025] Spatial transcriptome data of one case of triple-negative breast cancer was downloaded from GEO, annotated using RCTD, and the spatial expression of model genes was analyzed. The spatial expression results of the model genes are shown below. Figure 4 As shown, A and B are the dimensionality-reduced Umap maps of spatial transcriptomics data and the spatial distribution maps of each cluster; C and D are the annotated Umap maps and spatial distribution maps of each cell type; EG: spatial distribution of the 6 model genes. The inclusion of spatial transcriptomics data demonstrates that the risk score of this prognostic model can reveal "niche" or "functional unit" with clear spatial characteristics and poor prognosis within the triple-negative breast cancer tumor microenvironment. This model reflects the spatial biology of tumors and has clinical and biological significance.

[0026] Immunohistochemical staining was used to validate the expression of model genes in clinical samples from triple-negative breast cancer patients and their prognostic value. Tissue microarray samples from 48 triple-negative breast cancer patients at Wuhan University People's Hospital were used, along with complete clinicopathological and 5-year disease-free survival follow-up data. Protein expression of six model genes was detected using a standard procedure: paraffin sections were first dewaxed, hydrated, antigen-retrievaled, and blocked. Then, primary antibodies of specific concentrations (PINK1, PFKFB3, RSPO3, TREM-1, GZMB, and VEGFA at dilutions of 1:300, 1:300, 1:200, 1:200, 1:200, and 1:4000, respectively) were incubated overnight at 4°C, followed by incubation with HRP-labeled secondary antibody and DAB staining. Immunohistochemical scoring was determined by assessing staining intensity and the proportion of tumor cells with positive signals, calculated as: IHC score = intensity score × percentage score. The scoring criteria were defined as follows: percentage of positive cells, 0 (<10%), 1 (10–25%), 2 (26–50%), 3 (51–75%), 4 (>75%); staining intensity, 0 (no staining), 1 (light brown), 2 (brown), 3 (dark brown). Based on the total IHC score, triple-negative breast cancer samples were divided into positive and negative expression groups for each gene. For PINK1, PFKFB3, TREM1, GZMB, RSPO3, and VEGFA, cases with an IHC score greater than 4 were considered positive expression, and cases with a score equal to or less than 4 were considered negative expression.

[0027] The above experiments were used to clinically validate the model, directly verifying the protein expression of the six genes selected by bioinformatics in clinical samples and confirming their value as independent prognostic biomarkers. The results are as follows: Figure 5 As shown, the expression of PINK1, PFKFB3, TREM1, and VEGFA was significantly higher in relapsed patients than in non-relapsed patients. Conversely, the expression of GZMB and RSPO3 was significantly higher in non-relapsed patients than in relapsed patients. This indicates that GZMB and RSPO3 are protective proteins, and higher expression levels indicate a better prognosis; conversely, higher expression levels of PINK1, PFKFB3, TREM1, and VEGFA indicate a worse prognosis. Conversely, the model's risk score, RiskScore = (-0.158360246047182 × GZMB expression level) + (0.32235503219597 × PINK1 expression level) + (0.111243442274487 × PFKFB3 expression level) + (-0.0464626296260152 × RSPO3 expression level) + (0.0437384204335007 × TREM1 expression level) + (0.0364743584502037 × VEGFA expression level), indicates a higher risk score and a worse prognosis. In this score, GZMB and RSPO3 are negatively correlated with the score, meaning higher expression levels correspond to lower risk scores and better prognosis. The other four genes show the opposite trend. Furthermore, the conclusions drawn from this immunohistochemical analysis in clinical samples are consistent with the model score. Figure 5The results validated the model's prognostic ability. PINK1, PFKFB3, TREM1, and VEGFA were expressed significantly higher in relapsed patients than in non-relapsed patients. Conversely, GZMB and RSPO3 were expressed significantly higher in non-relapsed patients than in relapsed patients. This indicates that GZMB and RSPO3 are protective proteins, with higher expression levels indicating a better prognosis; conversely, higher expression levels of PINK1, PFKFB3, TREM1, and VEGFA indicate a worse prognosis. Conversely, the model's risk score, RiskScore = (-0.158360246047182 × GZMB expression level) + (0.32235503219597 × PINK1 expression level) + (0.111243442274487 × PFKFB3 expression level) + (-0.0464626296260152 × RSPO3 expression level) + (0.0437384204335007 × TREM1 expression level) + (0.0364743584502037 × VEGFA expression level), indicates a higher risk score and a worse prognosis. In this score, GZMB and RSPO3 are negatively correlated with the score, meaning higher expression levels correspond to lower risk scores and better prognosis. The other four genes show the opposite trend. Furthermore, the conclusions drawn from this immunohistochemical analysis in clinical samples are highly consistent with the model score. Figure 5 The results can verify the model's prognostic ability.

[0028] 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, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer, characterized in that, Includes the following steps: (1) Transcriptome data and prognostic data of triple-negative breast cancer patients and normal individuals were obtained from the database, and differentially expressed genes of triple-negative breast cancer were obtained by comparing transcriptome data. (2) The intersection of pyroptosis-related genes and the differentially expressed genes was obtained from the database. Based on the genes in the intersection, genes related to the prognosis of triple-negative breast cancer were screened by Lasso Cox regression analysis and used to construct a risk scoring model. The risk scoring model is expressed as follows: Risk Score=β1×Exp(mRNA1)+β2×Exp(mRNA2)+...+β n ×Exp(mRNAn) Where β1, β2, ..., β n is the coefficient of the gene, and mRNA1, mRNA2...mRNAn are the expression levels of the gene.

2. The method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer according to claim 1, characterized in that, The genes identified in step (2) that are associated with the prognosis of triple-negative breast cancer are: PINK1, GZMB, PFKFB3, RSPO3, TREM1 and VEGFA.

3. The method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer according to claim 1, characterized in that, The risk scoring model in step (2) is as follows: RiskScore = (-0.15836×GZMB expression level) + (0.32236×PINK1 expression level) + (0.11124×PFKFB3 expression level) + (-0.04646×RSPO3 expression level) + (0.04374×TREM1 expression level) + (0.03647×VEGFA expression level).

4. The method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer according to claim 1, characterized in that, In step (2), the prognostic model is evaluated based on the training set data and the test set data.

5. The method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer according to claim 4, characterized in that, The evaluation method is as follows: Calculate the risk score for each subject in the training dataset based on the risk scoring model; Scoring was analyzed using Kaplan-Meier curves from the training and test sets, as well as time-dependent receiver operating characteristic curves. The predictive efficacy of the model was evaluated using Kaplan-Meier curves from the training and test sets, as well as time-dependent receiver operating characteristic (ROC) curves.

6. The method for constructing a pyroptosis-related prognostic model for triple-negative breast cancer according to claim 5, characterized in that, Patients were divided into high-risk and low-risk groups based on the median value.

7. A pyroptosis-related prognostic model for triple-negative breast cancer, characterized in that, It is constructed by the method described in any one of claims 1-6.

8. The pyroptosis-related prognostic model for triple-negative breast cancer according to claim 7, characterized in that, The area under the ROC curve for the five-year survival prognosis of triple-negative breast cancer patients using the model reached 0.

83.

9. The application of the pyroptosis-related prognostic model for triple-negative breast cancer as described in claim 7 or 8 in assessing the prognosis of patients with triple-negative breast cancer.

10. The application according to claim 9, characterized in that, The expression levels of genes related to prognosis were measured in tumor samples from patients with triple-negative breast cancer.

Images