Immune regulatory mechanisms of M2 macrophage polarization and efferocytosis in diabetic kidney disease: an integrated screening study with therapeutic implications - PubMed (original) (raw)
Immune regulatory mechanisms of M2 macrophage polarization and efferocytosis in diabetic kidney disease: an integrated screening study with therapeutic implications
Yi Kang et al. Front Endocrinol (Lausanne). 2025.
Abstract
Background: The imbalance in macrophage phenotype transition is a central mechanism driving chronic inflammation in diabetic kidney disease (DKD). Macrophages can polarize toward the M2 phenotype via efferocytosis, exerting anti-inflammatory and pro-resolving effects. However, the identification and functional validation of regulatory genes governing M2 macrophage and efferocytosis in DKD remain to be thoroughly explored.
Methods: Differentially expressed genes were obtained based on GSE96804 and GSE30122 data sets. Based on efferocytosis-related genes (ERGs) and M2 polarization-related genes (MRGs), ERG and MRG scores were computed in the GSE96804 dataset. Weighted gene co-expression network analysis (WGCNA) was carried out to identify critical module genes. Finally, macrophage-efferocytosis-related DEGs (MEDEGs) were identified. Further, machine learning (ML)-support vector machine (SVM), BORUTA, and lasso regression-were employed to identify hub genes and build Nomogram predictive model. Additionally, hub genes were confirmed through animal experiments.
Results: A total of 35 MEDEGs were identified. ML recognized 3 hub genes-MCUR1, CYP27B1, and G6PC. Hub genes were notably downregulated in DKD group and exhibited high predictive ability. Furthermore, the Nomogram model based on key genes has shown potential in predicting DKD. The findings were further validated through transcriptome sequencing of DKD model.
Conclusion: This study uncovered 3 hub genes-MCUR1, CYP27B1, and G6PC-linked to M2 polarization, efferocytosis, and DKD. These genes may contribute to DKD pathogenesis, providing novel targets for early diagnosis and therapeutic interventions in DKD.
Keywords: M2 macrophage; bioinformatics; diabetic kidney disease; efferocytosis; immune regulatory.
Copyright © 2025 Kang, Jin, Zhou, Zheng, Li, Wang, Zhou, Lv and Wang.
Conflict of interest statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Figures
Figure 1
Research flowchart.
Figure 2
Efferocytosis and M2 macrophage polarization in DKD. (A, B) Volcano plot. (A) DEGs 1; (B) DEGs 2. (C, D) Heatmap. (C) DEGs 1; (D) DEGs 2; (E) Upset plot of intersecting DEGs. (F) MRGs scores (G) ERGs scores (* P < 0.05,**** P < 0.001).
Figure 3
Analysis of gene modules associated with M2 macrophage polarization and efferocytosis in DKD. (A, B) Clustering analysis of characteristic genes (A) M2 macrophages; (B) efferocytosis; (C) Selection of the soft-thresholding power; (D) WGCNA dendrogram: Co-expression modules identified and color-coded, with a total of 16 characteristic modules; (E, F) Heatmap of correlations between co-expression modules (E) M2 macrophage polarization; (F) efferocytosis.
Figure 4
Enrichment analysis of MEDEGs. (A) Venn diagram of DEGs, ERGs, and MRGs.; (B) PPI network; (C) GO enrichment analysis; (D) KEGG enrichment analysis.
Figure 5
ML selection of hub genes. (A) Five-fold cross-validation accuracy of the SVM method under different feature numbers; (B) Five-fold cross-validation error of the SVM method under different feature numbers; (C, D) BORUTA algorithm results. (E) Cross-validation curve of LASSO regression; (F) Sankey diagram integrating the three machine learning approaches.
Figure 6
Validation and clinical correlation analysis. **(A)**hub genes in GSE96804 dataset; (B) ROC curves of hub genes in GSE96804; (C) hub genes in the GSE30122 dataset; (D) hub genes in the Ju CKD TubInt dataset; (E) ROC curves of hub genes in Ju CKD TubInt; (F) Correlation between hub genes and clinical parameters in Ju CKD TubInt; (G) hub genes in the Ju CKD Glom dataset; (H) ROC curves of hub genes in Ju CKD Glom; (I) Correlation between hub genes and clinical parameters in Ju CKD Glom (* P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001).
Figure 7
Nomogram model. (A) Nomogram; (B) Calibration curve; (C) Risk scores in GSE96804 dataset; (D) ROC curve of risk scores in GSE96804 dataset; (E) Risk scores in Ju CKD TubInt dataset; (F) ROC curve of risk scores in Ju CKD TubInt dataset; (G) Correlation between the risk scores and Scr; (H) Correlation between the risk scores and GFR.
Figure 8
Immune infiltration. (A) Relative proportions of 22 immune cells; (B) Box plot comparing immune cell infiltration levels; (C) Heatmap illustrating the correlations between hub genes and immune cell types (* P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001).
Figure 9
Sc-RNA seq expression analysis (A) t-SNE plot of single-cell sequencing data; (B) Bubble plot of MCUR1 expression levels; (C) Bubble plot of CYP27B1 expression levels; (D) Bubble plot of G6PC expression levels.
Figure 10
Consensus clustering analysis. (A) Consensus clustering matrix (k = 2); (B) CDF curve; (C) Area under the CDF curve; (D) PCA plot of two subtypes; (E-G) Box plots of hub genes expression (* P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001).; (H) GSVA analysis between two subtypes.
Figure 11
Molecular regulatory network and drug prediction. (A) mRNA-TF network; (B) mRNA-miRNA network; (C) TF-miRNA-mRNA network; (D) Drug prediction.
Figure 12
Animal experiment validation. (A) HE, PAS, and Masson staining (×400; 50μm). (B) kW/BW at 20 weeks. (C) Scr at 20 weeks. (D) mA/uCr at 20 weeks. (E-G) mRNA expression of Mcur1, Cyp27b1, and G6pc (* P < 0.05, ** P < 0.01, *** P < 0.001).
References
- Federation ID . IDF Diabetes Atlas 2023. 10th ed. Brussels, Belgium: International Diabetes Federation. IDF Diabetes Atlas; (2023).
MeSH terms
LinkOut - more resources
Full Text Sources
Medical