Comprehensive identification of a disulfidptosis-associated long non-coding RNA signature to predict the prognosis and treatment options in ovarian cancer

Abstract

Purpose: Distinguished from cuproptosis and ferroptosis, disulfidptosis has been described as a newly discovered form of non-programmed cell death tightly associated with glucose metabolism. However, the prognostic profile of disulfidptosis-related lncRNAs (DRLRs) in ovarian cancer (OC) and their biological mechanisms need to be further elucidated.

Materials and methods: First, we downloaded the profiles of RNA transcriptome, clinical information for OC patients from the TCGA database. Generated from Cox regression analysis, prognostic lncRNAs were utilized to identify the risk signature by least absolute shrinkage and selection operator analysis. Then, we explored the intimate correlations between disulfidptosis and lncRNAs. What's more, we performed a series of systemic analyses to assess the robustness of the model and unravel its relationship with the immune microenvironment comprehensively.

Results: We identified two DRLR clusters, in which OC patients with low-risk scores exhibited a favorable prognosis, up-regulated immune cell infiltrations and enhanced sensitivity to immunotherapy. Furthermore, validation of the signature by clinical features and Cox analysis demonstrated remarkable consistency, suggesting the universal applicability of our model. It's worth noting that high-risk patients showed more positive responses to immune checkpoint inhibitors and potential chemotherapeutic drugs.

Conclusion: Our findings provided valuable insights into DRLRs in OC for the first time, which indicated an excellent clinical value in the selection of management strategies, spreading brilliant horizons into individualized therapy.

Keywords: disulfidptosis; immunotherapy; lncRNA; ovarian cancer; signature.

Figure 1

Generation of DRLRs in the model. (A) The Sankey diagram for all lncRNAs associated with disulfidptosis genes. (B) The forest plot for DRLRs with prognostic value. (C) The heat map showing correlations between disulfidptosis genes and core lncRNAs involved in the signature. Red represents positive correlation; Blue represents negative correlation.

Figure 2

Overview of the risk stratification in the model. (A-C) The distributions of the risk scores in the entire, training and testing cohort, respectively. (D-F) Heat maps for the expression divergence of the six lncRNAs between different risk samples in the whole, training, and testing sets, respectively. (G-I) The relationship between survival status and risk scores in (G) the whole cohort, (H) the training cohort and (I) the testing cohort.

Figure 3

Survival curves and the Cox regression analysis of the signature. (A-C) Kaplan-Meier plots of low- and high-risk patients in (A) TCGA, (B) the training group and (C) the testing group. (D, E) Forest plots of (D) univariate Cox regression and (E) multivariate Cox regression.

Figure 4

Evaluation and perfection of the prediction model. (A) Comparison between the risk score and the other clinicopathological variable in the accuracy of prognosis prediction. (B) Predicted survival rates on the basis of our signature at 1-, 3- and 5-year. (C) C-index curves of the risk score and other clinicopathological signatures. (D) The nomogram integrating grade, age and the risk score for survival outcomes prediction. (E) The calibration curve of the combination nomogram.

Figure 5

PCA results. Differences in distribution between the high- and low-risk groups in terms of all genes (A), disulfidptosis-related genes (B), disulfidptosis-related lncRNAs (C) and six hub lncRNAs (D).

Figure 6

Functional enrichment analyses. GSEA analysis of the (A) high- and (B) low-risk samples. (C) Column plots of GO terms. BP, biological processes; CC, cellular components; MF, molecular functions.

Figure 7

Immune microenvironment profiles in different risk samples. (A, B) The box plot of (A) immune cells and (B) immunological functions infiltration between different groups. (C) The stromal, immune and ESTIMATE scores in the low- and high-risk groups. (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 8

Tumor sensitivity to immunotherapy and chemotherapy. (A) Differences in immunoescape between low- and high-risk groups. (*p<0.05). (B) Sensitive chemotherapeutic agents for high-risk population.