Metformin mitigates SASP secretion and LPS-triggered hyper-inflammation in Doxorubicin-induced senescent endothelial cells

Abstract

Introduction: Doxorubicin (DOX), a chemotherapeutic drug, induces senescence and increases the secretion of senescence-associated secretory phenotype (SASP) in endothelial cells (ECs), which contributes to DOX-induced inflammaging. Metformin, an anti-diabetic drug, demonstrates senomorphic effects on different models of senescence. However, the effects of metformin on DOX-induced endothelial senescence have not been reported before. Senescent ECs exhibit a hyper-inflammatory response to lipopolysachharide (LPS). Therefore, in our current work, we identified the effects of metformin on DOX-induced endothelial senescence and LPS-induced hyper-inflammation in senescent ECs. Methods: ECs were treated with DOX ± metformin for 24 h followed by 72 h incubation without DOX to establish senescence. Effects of metformin on senescence markers expression, SA-β-gal activity, and SASP secretion were assessed. To delineate the molecular mechanisms, the effects of metformin on major signaling pathways were determined. The effect of LPS ± metformin was determined by stimulating both senescent and non-senescent ECs with LPS for an additional 24 h. Results: Metformin corrected DOX-induced upregulation of senescence markers and decreased the secretion of SASP factors and adhesion molecules. These effects were associated with a significant inhibition of the JNK and NF-κB pathway. A significant hyper-inflammatory response to LPS was observed in DOX-induced senescent ECs compared to non-senescent ECs. Metformin blunted LPS-induced upregulation of pro-inflammatory SASP factors. Conclusion: Our study demonstrates that metformin mitigates DOX-induced endothelial senescence phenotype and ameliorates the hyper-inflammatory response to LPS. These findings suggest that metformin may protect against DOX-induced vascular aging and endothelial dysfunction and ameliorate infection-induced hyper-inflammation in DOX-treated cancer survivors.

Keywords: LPS; SASP; doxorubicin; endothelial senescence; metformin; senomorphics.

FIGURE 1

Metformin inhibited DOX-induced upregulation of senescence markers and SA-β-gal activity in endothelial cells. (A) Schematic diagram of the experimental design. Both EA.hy926 endothelial derived cells and HUVECs were treated for 24 h with 0.5 µM DOX ± metformin (0.5–5 mM, added 24 h before DOX). Thereafter, DOX was removed and the cells were incubated in DOX-free media with or without metformin for an additional 72 h for protein expression experiments or 120 h for measurement of SA-β-gal staining. Expression levels of senescence markers including p-p53, p53, and p21 in EA.hy926 cells (B–D), and HUVECs [(E–G), respectively] were measured using western blot (n = 4–8). Representative images of western blots are shown. Values were normalized to α-tubulin and expressed relative to cells treated with DOX alone. (H) Images of SA-β-gal staining in control, DOX-treated, and DOX + metformin co-treated cells are shown in HUVECs. Images were analyzed and the percentage of SA-β-gal positive cells were calculated (n = 6–8). Values are presented as means ± SEM. Data were analyzed by one-way ANOVA followed by a Dunnet’s multiple comparisons test (Figures 1B,D,F–H) or non-parametric Kruskal–Wallis tests followed by Dunn’s post hoc test (Figures 1C, E); * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Schematic diagram created with BioRender.com.

FIGURE 2

Metformin decreased DOX-induced SASP factors in conditioned media of HUVECs. HUVECs were treated for 24 h with 0.5 µM DOX ± metformin (2 and 5 mM, added 24 h before DOX). Thereafter, DOX was removed and the cells were incubated in DOX-free media with or without metformin for an additional 72 h. Conditioned media were collected and the protein expression of SASP factors including IL-6, TNF-α, MIP-1α, MCP-3, MCP-1, CXCL1, CXCL2, IL-1 β, and IL-8 [(A–I), respectively (n = 5–8)] was determined by Luminex multiplex platform. Values were normalized to the protein concentration of the cells determined by BCA. Values are shown as means ± SEM. Data were analyzed by one-way ANOVA followed by a Dunnet’s multiple comparisons test (Figures 2A,B, D–E, G, I) or non-parametric Kruskal–Wallis tests followed by Dunn’s post hoc test (Figures 2C,F,H); * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

FIGURE 3

Metformin decreased the protein expression of SASP protease, MMP-3, and endothelial adhesion molecules in HUVECs. HUVECs were treated for 24 h with 0.5 µM DOX ± metformin (2 and 5 mM, added 24 h before DOX). Thereafter, DOX was removed and the cells were incubated in DOX-free media with or without metformin for an additional 72 h. The protein expression of the SASP protease MMP-3 was determined in (A) culture media by Luminex and (B) cell lysate by western blotting (n = 8–9). Expression of the endothelial adhesion molecules ICAM-1 in (C) culture media (n = 8) and (D) cell lysate (n = 4), and (E) E-selectin (n = 8) in the culture media were determined. Representative images of western blots are shown. Expression values in the media were normalized to the protein concentration of the cells determined by BCA. Values are shown as means ± SEM. Data were analyzed by one-way ANOVA followed by a Dunnet’s multiple comparisons test (Figures 3A,B,D–E) or non-parametric Kruskal–Wallis tests followed by Dunn’s post hoc test (Figure 3C); * p <0.05, ** p <0.01, *** p <0.001, **** p < 0.0001.

FIGURE 4

Signaling pathways associated with the protective effect of metformin on endothelial senescence in HUVECs. HUVECs were treated for 24 h with 0.5 µM DOX ± metformin (2 and 5 mM, added 24 h before DOX). Thereafter, DOX was removed and the cells were incubated in DOX-free media with or without metformin for an additional 72 h. Protein expressions of phospho- (A) JNK, (B) p38, (C) NF-κB p65, and (D) AMPK (n = 4–7) were quantified using western blot. Representative images of western blots are shown. Values were normalized to total protein or α-tubulin and expressed relative to cells treated with DOX alone. Expressed values are presented as mean ± SEM. Data were analyzed by one-way ANOVA followed by a Dunnet’s multiple comparisons test (Figure 4A,C) or non-parametric Kruskal–Wallis tests followed by Dunn’s post hoc test (Figure 4B,D); * p <0.05, *** p <0.001, **** p < 0.0001.

FIGURE 5

Metformin ameliorated LPS-triggered hyper-inflammation in DOX-induced senescent HUVECs. (A) Schematic diagram of the experimental design. HUVECs were treated for 24 h with 0.5 µM DOX ±5 mM metformin (added 24 h before DOX) or left untreated. Thereafter, DOX was removed and the cells were incubated in DOX-free media with or without metformin for an additional 72 h. The media were changed so that the assessed SASP factors in the media reflect only the effect of LPS. Then, cells were stimulated with LPS (30 ng/mL) for an additional 24 h. Thereafter, conditioned media were collected and the protein expression of SASP factors including (B) IL-6, (C) CXCL2, (D) MMP-3, (E) TNF-α, and (F) MCP-3 was determined by Luminex (n = 4). Values were normalized to the protein concentration of the cells determined by BCA. Expressed values are presented as mean ± SEM. Data were analyzed by two-way ANOVA followed by a Tukey multiple comparisons test. * compared to different treatment within the same group; * p <0.05, ** p <0.01, **** p < 0.0001. # compared to non-senescent cells with the same treatment; # p <0.05, #### p < 0.0001.

FIGURE 6

Metformin suppressed the expression of ICAM-1 and prevented NF-κB activation following LPS stimulation. HUVECs were treated for 24 h with 0.5 µM DOX ± 5 mM metformin (added 24 h before DOX) or left untreated. Thereafter DOX was removed and the cells were incubated in DOX-free media with or without metformin for an additional 72 h. Then, cells were stimulated with LPS (30 ng/mL) for an additional 24 h. (A) Representative images of western blot are shown and the expression levels of (B) ICAM-1, (C) p21, and (D) phospho-NF-κB p65 were measured (n = 4). Values were normalized to α-tubulin and expressed relative to DOX + LPS treated cells. Expressed values are presented as mean ± SEM. Data were analyzed by two-way ANOVA followed by a Tukey multiple comparisons test. * compared to different treatment within the same group; ** p <0.01, *** p <0.001, **** p < 0.0001. # compared to non-senescent cells with the same treatment; # p <0.05, ### p <0.001, #### p < 0.0001.