Fig. 4: Astragaloside IV attenuates diabetic kidney injury by preserving podocyte homeostasis.

a Integrated cellular landscape of therapeutic intervention. Uniform Manifold Approximation and Projection (UMAP) visualization of 151,009 high-quality cells from all experimental groups: Group I (non-diabetic db/m mice, n = 3), Group II (diabetic db/db mice, n = 3), and Group III (diabetic db/db + ASIV-treated, n = 3). Cells are colored by experimental condition. High-resolution clustering reveals ASIV-mediated preservation of podocyte integrity. b t-SNE and c UMAP visualizations of 1137 re-clustered podocytes from all experimental groups, with cells colored by treatment. Both visualizations show that ASIV treatment partially rescues the diabetes-induced transcriptional shift in podocytes. d Proportional abundance and quantitative assessment of podocyte across the three experimental groups. ASIV treatment increase the number of podocytes. The db/db mice exhibit significant podocyte loss compared to db/m controls, which is markedly attenuated by ASIV treatment. e Volcano plot for the differentially expressed genes in podocyte cells; Significantly regulated genes are highlighted, with key genes annotated. f Functional annotation of diabetes mice kidney-induced gene signatures whose diabetes-induced alterations are reversed by ASIV treatment. Bar plot showing the top significantly enriched phenotype ontology terms and the analysis identifies specific morphological and functional deficits in podocytes. g UMAP plots showing expression distribution of key genes involved in ferroptosis (GPX4, MAPK14, SLC7A11). h Quantitative validation of ASIV-mediated gene regulation. Bar chart comparing normalized expression levels (mean ± SEM; *p < 0.05, **p < 0.01) of ferroptosis-related markers (GPX4, MAPK14, SLC7A11) across experimental groups. ASIV treatment significantly counteracts diabetes-induced alterations in these critical regulators, supporting its role in modulating ferroptosis pathways in diabetic kidney disease. i Representative Western Blot bands analyzing key proteins under different treatments. Lanes display levels of phospho-p38 (p-p38), total p38, SLC7A11, GPX4, and LPO, with β-actin or GAPDH as loading controls. Treatments were: Control, Model (high glucose, HG), ASIV (HG + 30 μM ASIV), and C16-PAF (HG + ASIV + 5 μM C16-PAF). Data (mean ± SD; *P < 0.05, **P < 0.01) reveal that HG significantly upregulated p-p38 and LPO while downregulating SLC7A11 and GPX4, AS-IV treatment reversed these changes, and partially blocked by the p38 MAPK activator C16-PAF.