A Novel Cellular Senescence-related lncRNA Signature for Predicting the Prognosis of Breast Cancer Patients

Abstract

Background: Long non-coding RNA (lncRNA), a crucial regulator in breast cancer (BC) development, is intricately linked with cellular senescence. However, there is a lack of cellular senescence-related lncRNAs (CSRLs) signature to evaluate the prognosis of BC patients. Methods: Correlation analysis was conducted to identify lncRNAs associated with cellular senescence. Subsequently, a CSRL signature was crafted in the training cohort. The model's accuracy was evaluated through survival analysis and receiver operating characteristic curves. Furthermore, prognostic nomograms amalgamating cellular senescence and clinical characteristics were devised. Tumor microenvironment and checkpoint disparities were compared between low-risk and high-risk groups. The correlation between these signatures and treatment response in BC patients was also investigated. Finally, functional experiments were conducted for validation. Results: A signature comprising nine CSRLs was devised, which demonstrated adept prognostic capability in BC patients. Functional enrichment analysis revealed that tumor and immune-related pathways were predominantly enriched. Compared to the low-risk group, the high-risk group could benefit more from immunotherapy and certain chemotherapeutic agents. The expression of the 9 CSRLs was validated through in vitro experiments in different subtypes of BC cell lines and tissues. AC098484.1 was specifically verified for its association with senescence-associated secretory phenotypes. Conclusion: The CSRLs signature emerges as a promising prognostic biomarker for BC, with implications for immunological studies and treatment strategies. AC098484.1 has potential relevance in the treatment of BC cell senescence, and these findings improve the clinical treatment levels for BC patients.

Keywords: breast cancer; cellular senescence; drug therapy; immune microenvironment; lncRNA.

Figure 1

The flow chart of this study.

Figure 2

The construction of a prognostic signature in BC patients. (A) Forest plots showing the results of the univariate Cox regression analysis between the 64 CSRLs and OS of BC. (B) 24 CSRLs were selected by the LASSO regression model according to minimum criteria. (C) The coefficient of CSRLs were calculated by LASSO regression.

Figure 3

Screening of prognostic CSRLs in BC. (A) A prognostic co-expression network of the 9 CSR lncRNAs-mRNAs. (B) The Sankey diagram of the relationship between lncRNA and mRNA.

Figure 4

Prognosis values of the 9 CSRLs signatures in the train, test, and entire cohorts. The distribution of (A) risk scores, (B) survival time and survival status, (C) heat maps of 9 lncRNAs expression. (D) Kaplan-Meier survival curves of overall survival of BC patients between low-risk and high-risk groups in the train, test, and entire cohorts, respectively. (E) The AUC of the ROC curve shows the accuracy of the prognostic model in the train, test, and entire cohorts.

Figure 5

Construction of nomogram and calibration. (A, B) The hazard ratio (HR) and 95% confidence interval of risk score and clinical features were calculated using the univariate and multivariate Cox regression analysis. (C) Clinical prognostic nomogram was developed to predict 1-, 3-, and 5-year survival. (D) Time-dependent ROC curve analyses for predicting OS at 5 years by risk score age, stage, T stage (tumor size), M stage (distant metastasis), N stage. (E-G) Calibration curves showing nomogram predictions for 1-year, 3-year, and 5 year survival in the train, test, and all cohorts.

Figure 6

The results of functional analysis based on DEGs between low-risk and high-risk groups. (A, B) The enriched gene terms in GSEA. (C) Column diagrams of Gene Ontology analysis for the DEGs. (D) Column diagrams of KEGG analysis for for the DEGs.

Figure 7

TME, and checkpoint analysis in BC. (A) The box plots of immune cells between the low-risk and high-risk groups. (B) The box plots of immune related pathways between the low-risk and high-risk groups. (C) The box plots of checkpoint related genes between the low-risk and high-risk groups. ^Nsp ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 8

Immune score and drug sensitivities between low-risk and high-risk groups in BC. (A-C) Differences in ESTIMATE scores, immune scores, and stromal scores between the different risk score groups. Boxplots depict differences in estimated IC50 levels of (D) cisplatin, (E) doxorubicin, (F) etoposide, (G) gefitinib, (H) gemcitabine, and (I) paclitaxel between risk score groups.

Figure 9

(A-I) RT-PCR validation of 9 CSRLs using MCF-7, T47D, SK-BR-3, MDA-MB-231 human BC cells and MCF-10A human breast epithelial cells. And (J) RT-PCR validation of 9 CSRLs using human BC and paraneoplastic tissues. Expression of nine prognostic factors in BC: AC098484.1, LINC01235, LINC01871, EGOT, SEMA3B-AS1, AL358472.3, AP000851.2, MAPT-AS1, and LINC00987. *p < 0.05; **p < 0.01; ns, non-significant.

Figure 10

Knockdown of AC098484.1 promotes senescence in BC cells. (A) Reduced expression of AC098484.1 in MCF-7 and SK-BR-3 cells. (B) AC098484.1 was knocked down in MCF-7 cells, and its effect on cellular senescence was analyzed using SA-β-gal staining. (C) AC098484.1 was knocked down in MCF-7 cells, and its effect on cellular senescence was analyzed using SA-β-gal staining. *p < 0.05; **p < 0.01; ns, non-significant.