Identification of pyroptosis and hypoxia related molecular subtypes and a prognostic signature in colorectal cancer

Figures

• 

  Fig. 1. Flowchart of the overall study design and analytical workflow. The diagram summarizes data collection, identification of pyroptosis- and hypoxia-related genes (PHRGs), molecular subtype classification, prognostic model construction, immune and immunotherapy analyses, and experimental validation

• 

  Fig. 2. Landscape of pyroptosis- and hypoxia-related genes (PHRGs) in colorectal cancer (CRC). A Principal component analysis (PCA) based on PHRG expression profiles distinguishing CRC and adjacent normal tissues. B Differential expression of PHRGs between CRC and adjacent normal tissues. C Volcano plot and D heatmap illustrating PHRG expression differences. E Waterfall plot showing somatic mutation profiles of PHRGs in CRC. F Copy number variation (CNV) patterns of PHRGs across CRC samples

• 

  Fig. 3. Identification of pyroptosis- and hypoxia-related molecular subtypes based on PHRGs. A Correlation matrix of PHRG expression. B Cumulative distribution function (CDF) curves for consensus clustering. C Relative change in the area under the CDF curve. D Heatmap of the consensus matrix at the optimal cluster number. E Kaplan–Meier survival curves of the four PHRG-based molecular subtypes (PHRGclusters). F Expression patterns of PHRGs across different PHRGclusters. G Heatmap showing differential PHRG expression among PHRGclusters

• 

  Fig. 4. Functional characteristics and clinical associations of PHRG-related molecular subtypes. A Gene set variation analysis (GSVA) comparing pathway activities among PHRGclusters. B Distribution of sex across PHRGclusters. C Distribution of age across PHRG-related molecular subtypes. D Alluvial diagram illustrating relationships among tumour stage, PHRGclusters, and TNM_M status

• 

  Fig. 5. Construction of molecular subtypes and functional enrichment based on pyroptosis- and hypoxia-related differentially expressed genes (PHRDEGs). A Heatmap of the consensus matrix for PHRDEG-based clustering. B Kaplan–Meier survival curves of the three PHRDEGclusters. GO enrichment analysis of PHRDEGs in C biological process (BP), D cellular component (CC), and E molecular function (MF). F KEGG pathway enrichment analysis. G NF-κB signalling and H chemokine signalling pathway enrichment results. Distribution of PHRDEGclusters according to I TNM_M stage and J sex. K Alluvial diagram showing transitions among age, PHRDEGclusters, and tumour stage

• 

  Fig. 6. Construction and validation of the PHRDEG-related prognostic model and associated pathway enrichment. A Forest plot of univariate Cox regression analysis for PHRDEGs. B Cross-validation curve for least absolute shrinkage and selection operator (LASSO) regression. C Forest plot of multivariable Cox regression analysis identifying independent prognostic genes. Kaplan–Meier survival curves of different PHRDEGscore risk groups in the D TCGA, E GSE17536, and F GSE39582 cohorts. Gene set enrichment analysis (GSEA) of G angiogenesis, H interferon-gamma response, I allograft rejection, J inflammatory response, K IL6/JAK/STAT3 signalling, and L MYC targets V2 pathways

• 

  Fig. 7. Immune landscape associated with the PHRDEGscore. A Heatmap showing immune cell infiltration levels across different PHRDEGscore groups. B Correlation matrix among infiltrating immune cell types. C Differences in immune cell infiltration between high- and low-PHRDEGscore groups. D Correlations between immune cell infiltration and expression of prognostic model genes

• 

  Fig. 8. Exploratory analysis of the PHRDEGscore and immunotherapy response in a non-CRC cohort. A Kaplan–Meier survival curves of different PHRDEGscore groups in the IMvigor210 immunotherapy-treated cohort. B Comparison of immunotherapy response rates between PHRDEGscore groups. C Distribution of PHRDEGscore, survival status, and expression patterns of prognostic genes. Expression levels of immune checkpoint genes D CD96, E TNFSF4, F CD274, G TNFRSF9, H PDCD1LG2, and I CTLA4 in different PHRDEGscore groups

• 

  Fig. 9. Development of a nomogram model for survival prediction in colorectal cancer (CRC). Forest plots of A univariate and B multivariate Cox regression analyses incorporating clinical variables and the PHRDEGscore. C Nomogram integrating independent prognostic factors to predict survival probability. D Calibration curve evaluating the agreement between predicted and observed survival

• 

  Fig. 10. qRT-PCR validation of prognostic gene expression in CRC tissues. Relative mRNA expression levels of a TM4SF4, b OXCT1, c HES4, d GAL, e HLA-DQB2, and f SHROOM2 in 10 paired CRC tissues and adjacent non-cancerous tissues. Statistical significance is indicated as *p < 0.05; **p < 0.01; ***p < 0.001; NS, not significant

• 

  Fig. 11. Western blot validation of selected prognostic genes in CRC. Representative western blot images showing increased protein expression of HES4 and SHROOM2 in CRC tissues compared with paired adjacent normal tissues (n = 6 pairs), whereas OXCT1 did not show a significant difference. Actin was used as a loading control