Genetic variation profile of the PRGs in CC
First of all, we found that AIM2, CASP3, CASP4, CASP5, CASP6, CASP8, GSDMB, GSDMC, GZMB, IL18, NLRP2, NLRP7, NOD2, PYCARD, and TNF were significantly upregulated in CC tissues, while ELANE, NLRP1, NOD1, and PJVK were significantly downregulated (Fig 1A). Survival prognosis analysis showed that IL 1B and PRKACA were significantly correlated with the prognosis of CC patients. The patients with low expression of IL 1B and high expression of PRKACA have a better survival prognosis (Fig 2B/C). Fig 2D shows the location of CNV alterations on chromosomes in PRGs, and that CNV changes were prevalent in these genes, with PRKACA having the highest probability of CNV amplification and GPX4 having the highest frequency of CNV deletion (Fig 1E). We also analyzed the somatic mutations of these PRGs in CC and showed that the overall mutation frequency was not high (≤4%) (Fig 1F).
Construction and evaluation of the effectiveness of prognostic signature
We performed a survival analysis of PRGs in the training dataset using a univariate Cox proportional regression model and screened for eight PRGs with prognostic value (Fig 2A), of which GZMB and TNF were significantly highly expressed in CC tissue and NOD1 was significantly low (Supplementary Fig 1B). Next, the pyroptosis-related signature was constructed by further downscaling the prognosis-related PRGs through Lasso regression (Fig 2B-C), and the risk score was calculated for each patient, risk score==(-0.0037×EXPGPX4) + (0.0377×EXPGSDME) + (-0.0043×EXPGZMA) + (-0.026×EXPGZMB) +(0.0017×EXPIL1B) + (0.4741×EXPNOD1) + (-0.1298×EXPPRKACA) + (0.0642×EXPTNF). Based on the median, patients were divided into low-risk and high-risk groups, and survival prognostic analysis showed a significant survival benefit for patients in the low-risk group (P < 0.001, Fig 3A), its AUC was 0.789. In test dataset A and test dataset B, the difference in survival prognosis (P < 0.05, Fig 3B/C) and predictive efficacy (AUC were 0.707 and 0.943, respectively, Fig 3E/F) were similarly validated, confirming the overall accuracy and validity of the prognostic signature. Interestingly, this prognostic model showed a better prognosis for patients in the low-risk group than in the high-risk group in Glioblastoma multiforme (GBM), Brain Lower Grade Glioma (LGG), Liver hepatocellular carcinoma (LIHC), and Uterine Corpus Endometrial Carcinoma (UCEC) (Supplementary Fig 2).
Univariate and multivariate COX analyses combined with patient Age, Stage, Grade, showed that the pyroptosis-related signature was an independent prognostic factor for OS in CC patients (p < 0.001, Fig 4A/B). In addition, the risk score had the highest AUC value (AUC=0.794, Fig 4C) compared to other clinical features, and the DCA further demonstrated the clinical usefulness of the signature (Fig 4D). The Nomogram for both the signature and clinical features is stable and accurate and can be used to predict 1-year, 3-year, and 5-year survival rates in CC patients (Fig 4E).
Correlation of TMB, TME, and Immune cell infiltration with the prognostic signature
Previous studies have reported an important role for the TME in the occurrence and development of tumors(21). Given that TMB predicts patient response to immunotherapy, our study found that patients in the high-risk group were accompanied by low TMB scores, and the risk score was significantly negatively correlated with TMB score. The results of the survival analysis showed that patients with high-risk scores combined with low TMB scores showed a significant survival advantage (Fig 7A-C). Compared to the high-risk group, ESTIMATEScore, ImmuneScore, and StromalScore were all higher in the low-risk group (Fig 7D).
The relationship between immune response and risk score based on the TIMER, QUANTISEQ, MCPCOUNTER, and CIBERSORT algorithms is shown in Fig 5E. The risk score was negatively related to the infiltration level of B cell, T cell CD8+, Macrophages, and positively to the infiltration level of Neutrophils. The analysis of immune function differences showed that antigen-presenting cell (APC) co-inhibition/ APC_co_stimulation, CC chemokine receptor (CCR), Check-point, Cytolytic_activity, human leukocyte antigen (HLA), Inflammation promoting, T cell co inhibition/ stimulation were significantly more active in the low-risk group (Fig 6 A).
In the view of the above analysis of immune function, differences showed significant differences in HLA between the two groups, HLA-related gene expression levels were further analyzed as shown in Fig 6 B. HLA-DMA, HLA-DQB1, HLA HLA-DMA, HLA-DQB1, HLA-DQB1, HLA-DRA, etc. were significantly higher in the low-risk group. Due to the importance of immunotherapy with checkpoint inhibitors, we further analyzed the differences in immune checkpoint expression between the two groups and showed significant differences in the expression of PDCD1LG2, TMIGD2, CD27, CD40LG, etc. between the two groups (Fig 6C). The latest research suggests that m6A can regulate immune response and participate in the regulation of the tumor immune microenvironment(22). Comparison of m6A-related mRNA expression between high and low-risk groups showed that YTHDF3, LRPPRC were significantly more highly expressed in the high-risk group (Fig 6D).
Construction of core gene ceRNA interoperability network
The interrelationships between the 35 PRGs included in this study were retrieved from the STRING database and a PPI network was constructed with a composite score of >0. 4. A close and complex network of interrelationships was found between these genes (Supplementary Fig 3A). And count the degree of connectivity between these genes (Supplementary Fig 3B). We further analyzed the relationship between mRNA expression levels in the pyroptosis-related signature and CC Grade staging and found that GPX4, GZMA, and GZMB expression was significantly higher in the G3-4 group compared to the G1-2 group (P < 0.05, Supplementary Fig 3C/E/F). The connectivity of GPX4, GZMA, and GZMB in the above PPI network was 4, 4, and 10, respectively, suggesting that GZMB may be involved in the tumor progression of CC.
Further pan-cancer expression analysis was performed for GZMB, showing significant high expression of GZMB in Cholangiocarcinoma (CHOL), Colon adenocarcinoma (COAD), Esophageal carcinoma (ESCA), Glioblastoma multiforme (GBM), Head and Neck squamous cell carcinoma (HNSC), Kidney renal clear cell carcinoma (KIRC), Rectum adenocarcinoma (READ), Stomach adenocarcinoma (STAD), and Uterine Corpus Endometrial Carcinoma (UCEC) tissues and significant low expression in Lung adenocarcinoma (LUAD) and Lung squamous cell carcinoma (LUSC) tissues (Fig 7A). And prognostic correlation analysis showed that GZMB was negatively correlated with survival in Bladder Urothelial Carcinoma (BLCA), Breast invasive carcinoma (BRCA), Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), Ovarian serous cystadenocarcinoma (OV), Skin Cutaneous Melanoma (SKCM), UCEC, and Uveal Melanoma (UVM), while it was positively correlated with GBM, KIRC, Kidney renal papillary cell carcinoma (KIRP), Acute Myeloid Leukemia (LAML), and Brain Lower Grade Glioma (LGG) (P<0.05, Fig 7B). To clarify the potential molecular regulation mechanism of GZMB in CC, we further constructed an mRNA-miRNA-LncRNA interaction network. First, based on the comprehensive prediction of TarBase and mirTarBase databases, we screened out miRNAs that can target and regulate the core gene GZMB, and subsequently, 10 miRNAs were obtained by taking the intersection (Fig 7C). Further analysis showed that miR-378a was significantly upregulated in CC (Fig 7D) and its high expression patients have a better prognosis for survival (Fig 7E). The LncRNA targets upstream of miR-378a were obtained from the LncBase and StaBase databases, and a total of 5 LncRNAs (Fig 7F) were obtained by taking the intersection to finally construct the mRNA-miRNA-LncRNA interactions network (Fig 7I). Next, these five LncRNAs were also subjected to differential expression analysis and survival prognosis analysis, which showed that TRIM52-AS1 was significantly downregulated in CC (Fig 7G), and its high expression level showed a significant survival advantage (Fig 7H). Thus, the GZMB - miR-378a - TRIM52-AS1 regulatory axis may play an important regulatory role in the development of CC.