Construction of a novel necroptosis-related lncRNA signature for prognosis prediction in esophageal cancer

Abstract

Background: Esophageal cancer (EC), one highly malignant gastrointestinal cancer, is the 6th leading cause of cancer-related deaths worldwide. Necroptosis and long non-coding RNA (lncRNA) play important roles in the occurrence and development of EC, but the research on the role of necroptosis-related lncRNA in EC is not conclusive. This study aims to use bioinformatics to investigate the prognostic value of necroptosis-related lncRNA in EC.

Methods: Transcriptome data containing EC and normal samples, and clinical information were obtained from the Cancer Genome Atlas database. 102 necroptosis-related genes were obtained from Kanehisa Laboratories. Necroptosis-related lncRNAs were screened out via univariate, multivariate Cox and the least absolute shrinkage and selection operator regression analyses to construct the risk predictive model. The reliability of the risk model was evaluated mainly through quantitative real-time PCR (qRT-PCR), the receiver operating characteristic (ROC) curves and the constructed nomogram. KEGG pathways were explored in the high- and low-risk groups of EC patients via gene set enrichment analyses (GSEA) software. Immune microenvironment and potential therapeutic agents in risk groups were also analyzed.

Results: A 6 necroptosis-related lncRNAs risk model composed of AC022211.2, Z94721.1, AC007991.2, SAMD12-AS1, AL035461.2 and AC051619.4 was established to predict the prognosis level of EC patients. qRT-PCR analysis showed upregulated Z94721.1 and AL035461.2 mRNA levels and downregulated AC051619.4 mRNA level in EC tissues compared with normal tissues. According to clinical characteristics, the patients in the high-risk group had a shorter overall survival than the low-risk group. The ROC curve and nomogram confirmed this model as one independent and predominant predictor. GSEA analysis showed metabolic and immune-related pathways enriched in the risk model. Most of the immune cells and immune checkpoints were positively correlated with the risk model, mainly active in the high-risk group. For the prediction of potential therapeutic drugs, 16 compounds in the high-risk group and 2 compounds in the low-risk group exhibited higher sensitivity.

Conclusions: Our results supported the necroptosis-related lncRNA signature could independently predict prognosis of EC patients, and provided theoretical basis for improving the clinical treatment of EC.

Keywords: EC; Immune; Necroptosis; Prognosis; lncRNA.

Fig. 1

The flow diagram of the study

Fig. 2

Screening necroptosis-related lncRNAs in EC. A 100 differentially expressed necroptosis-related lncRNAs in heat map. B The upregulated and downregulated necroptosis-related lncRNAs in volcano plot. C The network between necroptosis-related genes and lncRNAs

Fig. 3

Identification of necroptosis-related lncRNAs associated with overall survival. A 24 necroptosis-related lncRNAs relating to overall survival of EC patients selected by univariate Cox regression analysis in forest plot. B The different expressions of these 24 lncRNAs in tumor and normal tissues. C The tenfold cross-validation for variable selection in the Lasso model. D The Lasso coefficient profile of 13 necroptosis-related lncRNAs. E The correlation between necroptosis-related genes and lncRNAs in the Sankey diagram

Fig. 4

Prognosis values of the predictive signature composed of 6 necroptosis-related lncRNAs and qRT-PCR analysis. Kaplan–Meier survival curves of EC patients (A), the distribution of risk scores (B), survival time and survival status (C), heat maps of 6 necroptosis-related lncRNA expressions (D) between the high- and low-risk groups in the train (N = 80), test (N = 79), and entire (N = 159) sets. E qRT-PCR to detect mRNA levels of Z94721.1, AL035461.2, and AC051619.4 in normal and EC tissues. **p < 0.01, ***p < 0.001, all n = 5

Fig. 5

Kaplan–Meier survival curves of the risk groups sorted by different clinical features including age (A), gender (B), tumor stage (C) and tumor grade (D)

Fig. 6

Validation of the risk signature as an independent predictor of prognosis. A and B Univariate and multivariate Cox regression analyses of clinical features and the risk score with overall survival in the train, test and entire sets. C Using the ROC curves to assess the prediction performance of the risk signature for the 1-, 2- and 3-year overall survival of EC patients. AUC, the area under the ROC curve. D Comparison of the prediction accuracy of the risk signature and clinical factors, such as age, gender, grade and stage

Fig. 7

Construction of nomogram and calibration. A A nomogram for predicting the 1-, 2- and 3-year survival incidences of EC patients according to the risk score and clinical factors containing age, gender, grade and stage. B The 1-, 2- and 3-year calibration plot for proving the reliability of the nomogram. OS, overall survival

Fig. 8

Pathway enrichment analysis via GSEA software. A–C The top 5 KEGG pathways enriched in the high- and low-risk groups. NES, normalized enrichment score. NOM, nominal

Fig. 9

Investigation of immune microenvironment in the high- and low-risk groups. A and B The correlation between immune cells and the risk groups in a bubble chart. C Comparison of immune score between the high- and low-risk groups

Fig. 10

Investigation of tumor immunotherapy. A 6 immune functions significantly activated in the high-risk group. B The differences of immune checkpoint expressions between the high- and low-risk groups. C the prediction of potential therapeutic compounds for the risk model