Luteolin Alleviates Vascular Senescence Through Retinoic Acid–Peroxisome Proliferator-Activated Receptor Signaling and Lipid Metabolism Remodeling Combined with Multi-Omics Analysis

Figures

• 

  Scheme 1. Experimental design and study overview. This schematic diagram illustrates the overall experimental approach, including the intervention with luteolin (Lut) and the different experimental models used in this study.

• 

  Figure 1. Effects of Lut on cell viability and senescence in vascular endothelial cells. (A) Dose–response analysis of Lut (0.04–80 µM) on cell viability in canine vascular endothelial cells (cVECs) and feline vascular endothelial cells (fVECs), assessed by CCK-8 assay (n ≥ 12). (B) Cell viability after treatment with different concentrations of doxorubicin (Dox; 0.1–25.6 µM) in cVECs and fVECs (n ≥ 12). (C) Protective effects of Lut (0.625–2.5 µM) against Dox-induced cytotoxicity (0.8 µM in cVECs; 1.6 µM in fVECs) as measured by cell viability (n ≥ 12). (D) Senescence-associated β-galactosidase (SA-β-gal) activity in cVECs and fVECs under Dox treatment with or without Lut co-treatment (n ≥ 10). (E) Representative SA-β-gal staining images showing cellular senescence in NC, Dox, and Dox + Lut (0.625 µM) groups in both cVECs and fVECs (n ≥ 4). Scale bar = 100 μm. Data are shown as mean ± SD. Distinct lowercase letters above the bars denote significant differences among groups (p < 0.05).

• 

  Figure 2. Effects of Lut on oxidative stress and energy metabolism in vascular endothelial cells. cVECs and fVECs were treated with Dox (0.8 µM for cVECs; 1.6 µM for fVECs) with or without 0.625 µM Lut for 24 h. (A) Levels of inflammatory SASP markers in culture supernatants, including interleukin (IL)-1β, tumor necrosis factor (TNF)-α, IL-6, and IL-8. (B) Levels of matrix metalloproteinase (MMP)-1 and MMP-3. (C) Monocyte chemoattractant protein-1 (MCP-1) levels in supernatants and malondialdehyde (MDA) content in cell lysates. (D) Secretion of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF). (E) Intracellular superoxide dismutase (SOD) activity and nicotinamide adenine dinucleotide (NAD+) levels. NC: untreated control; Dox: Doxorubicin only; Dox + Lut: Dox combined with 0.625 µM Lut. Data are presented as mean ± SD (n ≥ 10). The numerical values in the middle of each bar denote the mean of replicates. Distinct lowercase letters above the bars denote significant differences among groups (p < 0.05).

• 

  Figure 3. Effects of Lut on physiological, behavioral, and serum parameters in senescence-accelerated SAMP8 mice. Young control (YC): 1-month-old non-senescent SAMP8 mice; old control (OC): 7-month-old aged SAMP8 mice; OC + Lut: 7-month-old SAMP8 mice treated with Lut (10 mg/kg, gavage for 8 weeks). Male mice, n = 10 per group. (A) Body weight over the gavage period. (B) Rotarod performance: latency to fall measured weekly from week 4 of gavage. (C) Balance beam test: time to cross the beam measured weekly from week 4 of gavage. (D) Terminal physiological parameters, including systolic/diastolic blood pressure (SBP/DBP), respiratory rate (breaths/min per body weight), heart rate (bpm), and blood oxygen saturation (%). (E) Terminal organ weights: heart, liver, lung, and kidney. (F–M) Serum levels of lipofuscin (F), IL-6 (G), VEGF (H), bFGF (I), MMP-1 (J), MMP-3 (K), NAD+ (L), and GPX (M). Data are presented as mean ± SD. The numerical values in the middle of each bar denote the mean of replicates. Distinct lowercase letters above the bars denote significant differences among groups (p < 0.05). In panels B and C, *** indicate statistically significant differences compared with the YC and OC…

• 

  Figure 4. Histological evaluation of the abdominal aortas, heart and liver in aged SAMP8 mice after Lut intervention. (A) Representative hematoxylin and eosin (HE)-stained sections of abdominal aortas under 10×, 50.4×, and 80× magnification. (B) Peak vascular tensile force of abdominal aortas (n = 5 per group). (C) Tunica media thickness of abdominal aorta walls (n = 3 per group, more than 10 points were sampled for each tissue). (D) Representative HE staining of heart sections at 2×, 10×, and 40× magnifications. Red circles indicate areas of inflammatory cell infiltration. (E) Representative HE staining of liver sections at 10×, 30×, and 80× magnifications. Arrows indicate areas of inflammatory cell infiltration. YC, young control; OC, aged control; OC + LUT, aged SAMP8 mice treated with Lut for 8 weeks. Data are shown as mean ± SD. The numerical values in the middle of each bar denote the mean of replicates. Distinct lowercase letters above the bars denote significant differences among groups (p < 0.05).

• 

  Figure 5. Transcriptomic analysis revealed Lut-induced activation of metabolic and signaling pathways in Dox-treated cVECs. (A) KEGG enrichment bubble plot of core pathways regulated by Lut in Dox-induced cVECs (Dox_Lut vs. Dox) based on GSEA. (B) GSEA enrichment plots of representative upregulated pathways: cAMP signaling, regulation of lipolysis in adipocytes, and glycerolipid metabolism. (C) Hierarchical clustering heatmap of core DEGs in the cAMP pathway. (D) Heatmap of shared core DEGs involved in both regulation of lipolysis and glycerolipid metabolism. (E,F) KEGG pathway maps of (E) cAMP signaling and (F) regulation of lipolysis, with significantly upregulated genes marked in red boxes and downregulated in green.

• 

  Figure 6. Lut (0.625 μM) activated retinol metabolism and PPAR signaling pathways in Dox-induced (0.8 μM) senescent cVECs. (A) Heatmap of representative DEGs in the retinol metabolism pathway comparing Lut + Dox and Dox groups. Expression values were normalized and Z-score-transformed. (B) Heatmap of core DEGs within the PPAR signaling pathway, showing transcriptional activation following Lut treatment. (C) KEGG pathway maps of retinol metabolism and PPAR signaling. Genes upregulated are marked in red and those downregulated in green, reflecting Lut-induced activation of lipid regulatory pathways in stressed endothelial cells.

• 

  Figure 7. Metabolomic analysis of cVECs treated with 0.625 μM Lut under 0.8 μM Dox-induced senescence. (A) PCA displaying clear separation of metabolic profiles between NC, Dox, and Dox + Lut groups (n = 4). QC samples were generated by combining equal quantities from all experimental samples, with tight clustering demonstrating good reproducibility and stability. (B) Volcano plot of differential metabolites in Dox + Lut vs. Dox. Red and blue dots represent significantly upregulated and downregulated metabolites, respectively (p < 0.05). (C) KEGG functional classification of differential metabolites, highlighting the dominant enrichment in lipid metabolism (19.69%). (D) KEGG pathway enrichment bubble plot of altered metabolites between Dox + Lut and Dox groups. Bubble size reflects pathway metabolite count; color indicates p value. (E) Log2 fold change in representative metabolites involved in retinol metabolism, including significantly increased retinoic acid and 9-cis-retinal, supporting retinoid activation. (F) Differential expression of lipid metabolism-related metabolites, including PUFAs (e.g., EPA and docosadienoic acid), prostaglandins, and steroid hormones (e.g., estriol and…

• 

  Figure 8. Transcriptomic activation of retinol and PPAR signaling pathways in aged SAMP8 aorta after 8-week oral administration of Lut. (A) Heatmap of DEGs in the retinol metabolism between OC_Lut and OC groups. (B) Heatmap of DEGs in the PPAR signaling pathway between OC_Lut and OC groups. (C) KEGG pathway annotation of PPAR signaling showing upregulated (red border) and downregulated (green border) genes. Key targets include PPARα, PPARδ, PPARγ and RXR heterodimers, with downstream targets involved in lipid metabolism (fatty acid oxidation, transport, and biosynthesis), bile acid metabolism, glycerophospholipid metabolism, adipocyte differentiation, and adaptive thermogenesis.