DNA Methylation Patterns of 5mC Regulators: Insights into Immune Microenvironment Regulation in Heart Failure

Figures

• 

  Fig. 1. The overview of the study

• 

  Fig. 2. Expression of 5mC DNA regulators. The boxplot of transcriptome expression status of 20 5mC regulators in training cohort (A). B Correlations among 20 5mC regulators in HF. C The enrichment results of GO KEGG analysis

• 

  Fig. 3. Screening key biomarkers and establishing a diagnostic model of HF by three machine-learning algorithms. A Random forest model. B SVM-RFE model. C and D LASSO regression analysis. (E) Veen diagram shows 8 key biomarkers among 20 5mC regulators. F Nomogram of the eight-gene 5mC regulators. G Calibration curve. H DCA plots. I The boxplot presents consistent expressions of four key regulators (MBD2, DNMT3B, ZBTB4 and UNG) in the validation cohort and the training cohort. J-Q Predicting value of diagnostic models for both training cohort and validation cohorts. ROC curves unveil high AUC values in hub genes

• 

  Fig. 4. Relationship between 5mC markers and LVEF in HF individuals

• 

  Fig. 5. Validation of four key 5mC regulators. A Masson staining shows the collagen deposits characteristic of fibrosis 4 weeks after surgery, which is significantly different from the control group. (Scale bars = 2000–100 μm). The semiquantitative analysis results of Masson’s staining by ImageJ. B-E The qPCR results of DNMT3B, MBD2, ZBTB4 and UNG. F WB results of DNMT3B and MBD2. G IF results of DNMT3B and MBD2. (Scale bars = 50 μm). n = 6, *p < 0.05, **p < 0.01, VS. the control group

• 

  Fig. 6. The ssGSEA results of key regulators in 5mC DNA modification among HF samples. A-B DNMT3B; C-D MBD2

• 

  Fig. 7. The 5mC associated HF IME. A IME cell compositions in HF and healthy samples. B Relationships between 5mc regulators and 22 immune cells. The links of different 5mC regulators with NK cells activated; C NTHL1; D ZBTB38; E TET3; and (F) UHRF2

• 

  Fig. 8. Two 5mC modification subtypes in HF samples using 15 DEGs. A Consensus clustering model with cumulative distribution function for k = 2–10. k means cluster count. B Consensus matrix heatmap defining two subtypes (k = 2) and their correlation area. C The consensus cluster of items (in column) at k = 2–10 (in row). D The delta area under the cumulative distribution function curve. E PCA shows remarkable differences in transcriptomes. F 5mC subtypes present distinct expression of the 15 DEGs

• 

  Fig. 9. Bioinformatic analysis of DEGs in two 5mC patterns. A The volcano plot shows DEGs in two patterns. B GO analysis based on DEGs. C-D GSEA analysis

• 

  Fig. 10. IME characteristics between two subtypes. Infiltrating immune cells show distinct abundance in 5mC modification clusters by IME (A) and ssGSEA (B). C ICP analysis. D The difference in HLA expression