A comprehensive pan-cancer analysis of necroptosis molecules in four gynecologic cancers

Abstract

Background: In recent years, it has been proved that necroptosis plays an important role in the occurrence, development, invasion, metastasis and drug resistance of malignant tumors. Hence, further evaluation and targeting of necroptosis may be of clinical benefit for gynecologic cancers (GCs).

Methods: To compare consistency and difference, we explored the expression pattern and prognostic value of necroptosis-related genes (NRGs) in pan-GC analysis through Linear regression and Empirical Bayesian, Univariate Cox analysis, and public databases from TCGA and Genotype-Tissue Expression (GTEx), including CESC, OV, UCEC, and UCS. We explored the copy number variation (CNV), methylation level and enrichment pathways of NRGs in the four GCs. Based on LASSO Cox regression analysis or principal component analysis, we established the prognostic NRG-signature or necroptosis-score for the four GCs. In addition, we predicted and compared functional pathways, tumor mutational burden (TMB), somatic mutation features, immunity status, immunotherapy, chemotherapeutic drug sensitivity of the NRG-signature based on NRGs. We also examined the expression level of several NRGs in OV samples that we collected using Quantitative Real-time PCR.

Results: We confirmed the presence of NRGs in expression, prognosis, CNV, and methylation for four GCs, thus comparing the consistency and difference among the four GCs. The prognosis and independent prognostic value of the risk signatures based on NRGs were determined. Through the results of subclass mapping, we found that GC patients with lower risk score may be more sensitive to PDL1 response and more sensitive to immune checkpoint blockade therapy. Drug susceptibility analysis showed that, 51, 45, 64, and 29 drugs with differences between risk groups were yielded in CESC, OV, UCEC, and UCS respectively. For OV, the expression differences of several NRGs in the tissues we collected were similar to that in TCGA.

Conclusion: Our comprehensive analysis of NRGs and NRG-signature demonstrated their similarity and difference, as well as their potential roles in prognosis and could guide therapeutic strategies, thus improving the outcome of GC patients.

Keywords: Gynecologic cancer; Immunity; Immunotherapy; Necroptosis; Prognosis.

Fig. 1

Expression variation and mutation frequency of filtered necroptosis-related genes (NRGs). A Expression levels of mutual differentially expressed NRGs in the four gynecological cancers (GCs). The color of the dots represents the degree of variance. Redder dots represent higher expression in cancer tissue. Bluer dots represent higher expression in normal tissue. The size of the bubbles indicates the adjusted P-value. Larger bubbles represent a lower adjusted P-value. The NRGs with adjusted P-value < 0.05 & |logFC|> 0.5 and NRGs that were significantly differentially expressed in all the four GCs were retained to produce the figure. B-E Top 10 NRGs with mutation rates in patients with CESC (B), OV (C), UCEC (D), and UCS (E). The small figure above shows the TMB, the number on the right shows the mutation frequency of each NRG, and the figure on the right shows the proportion of each variant

Fig. 2

GO enrichment analysis of differentially expressed necroptosis-related genes (NRGs) in the four gynecological cancers (GCs). A-D GO enrichment analysis for CESC (A), OV (B), UCEC (C), and UCS (D). FC represents fold change. Blue dots indicate genes that were downregulated in the GCs, and red dots indicate genes that were upregulated in the GCs. The size of the z-score is shown by the color

Fig. 3

KEGG enrichment analysis and PPI analysis of necroptosis-related genes (NRGs) in the four gynecological cancers (GCs). A-D ClueGO results of KEGG analysis of NRGs for CESC (A), OV (B), UCEC (C), and UCS D. The size of the dots indicates the number of genes attributed to the category. E–H The top 10 hub genes were selected by degree to establish PPI network for CESC (E), OV (F), UCEC (G), and UCS (H)

Fig. 4

Principal component analysis for necroptosis-related genes (NRGs). A Pearson correlation between CNV and NRGs expression level. The bubble color indicates the degree of correlation index. The bubble size indicates the FDR. B Spearman correlation between methylation of the NRGs and their expression. The bubble color indicates the degree of correlation index. The bubble size indicates the P-value. C-E PCA analysis was conducted among all (C), normal (D), and tumor samples (E) respectively. Different gynecological cancers (GCs) are shown in different colors

Fig. 5

Construction of the prognostic signature based on the optimal NRGs. A-D Kaplan–Meier survival curves show survival probability of high-risk or low-risk for CESC (A), OV (B), UCEC (C), and UCS (D) in training sets. E–H Kaplan–Meier survival curves show survival probability of high-risk or low-risk for CESC (E), OV (F), UCEC (G), and UCS (H) in validation sets. I-L Kaplan–Meier survival curves show survival probability of high-risk or low-risk for CESC (I), OV (J), UCEC (K), and UCS (L) in total sets. The blue curve represents patients in the low-risk group, and the red curve represents patients in the high-risk group

Fig. 6

Clinical value of risk score by independent prognostic analysis. A-H The Univariate Cox regression analysis and Multivariate Cox regression analysis for CESC (A-B), OV (C-D), UCEC (E–F), and UCS (G-H). We reduced the clinicopathological parameters to the ones all GCs shared (Age, Stage)

Fig. 7

Evaluation of HALLMARK pathways between the two risk groups. A-D The bar plots indicate the distribution of the t values of the GSVA scores calculated for several pathways for CESC (A), OV (B), UCEC (C), and UCS (D). The blue bars represent HALLMARK pathways that are upregulated in the high-risk group, and the green bars represent HALLMARK pathways that are downregulated in the high-risk group

Fig. 8

Evaluation of immune activity between the two risk groups. A-D The distribution of 22 different immune cells between high and low risk groups for CESC (A), OV (B), UCEC (C), and UCS (D). The blue violins represent the low-risk group, and the red violins represent the high-risk group

Fig. 9

Evaluation of immunotherapy between the two risk groups. A-D The subclass mapping for CESC (A), OV (B), UCEC (C), and UCS D. E–H The ICB response rates for CESC (E), OV (F), UCEC (G), and UCS H

Fig. 10

Evaluation of estimate-related scores between the low-risk and high-risk groups. A-D Comparison of Estimate Score (A), Immune Score (B), Stromal Score (C) and Tumor Purity (D) for CESC. E–H Comparison of Estimate Score (E), Immune Score (F), Stromal Score (G) and Tumor Purity (H) for OV. I-L Comparison of Estimate Score (I), Immune Score (J), Stromal Score (K) and Tumor Purity (L) for UCEC. M-P Comparison of Estimate Score (M), Immune Score (N), Stromal Score (O) and Tumor Purity (P) for UCEC. ∗ P < 0.05; ∗  ∗ P < 0.01; ∗  ∗  ∗ P < 0.001; ns: not significant. The blue violins represent the low-risk group, and the red violins represent the high-risk group

Fig. 11

Analysis of chemotherapeutic sensitivity based on the NRG-signature. A-H Relationships between risk scores and IC50 level of drugs. Only the drug with the most significant IC50 difference in each tumor was shown. A-D Drugs with IC50 values most significantly higher in the high-risk group for CESC (A), OV (B), UCEC (C), and UCS (D). E–H Drugs with IC50 values most significantly higher in the low-risk group for CESC (E), OV (F), UCEC (G), and UCS (H). The blue bars represent the low-risk group, and the red bars represent the high-risk group