Metformin suppresses vascular smooth muscle cell senescence by promoting autophagic flux

Abstract

Introduction: Vascular smooth muscle cell (VSMC) senescence in the vasculature results in vascular aging as well as age-related diseases, while metformin improves the inflamm-aging profile by enhancing autophagy. However, metformin's impact on VSMC senescence is largely undefined.

Objectives: To test the hypothesis that metformin exerts an anti-senescence role by restoring autophagic activity in VSMCs and vascular tissues.

Methods: Animal models established by angiotensin II (Ang II) induction and physiological aging and senescent primary VSMCs from the aortas of elderly patients were treated with metformin. Cellular and vascular senescence were assessed by measuring the amounts of senescence-associated β-galactosidase and senescence markers, including p21 and p53. Autophagy levels were assessed by autophagy-related protein expression, transmission electron microscope, and autolysosome staining. In order to explore the underlying mechanism of the anti-senescence effects of metformin, 4D label-free quantitative proteomics and bioinformatic analyses were conducted, with subsequent experiments validating these findings.

Results: Ang II-dependent senescence was suppressed by metformin in VSMCs and vascular tissues. Metformin also significantly improved arterial stiffness and alleviated structural changes in aged arteries, reduced senescence-associated secretory phenotype (SASP), and improved proliferation and migration of senescent VSMCs. Mechanistically, the proteomic analysis indicated that autophagy might contribute to metformin's anti-senescence effects. Reduced autophagic flux was observed in Ang II-induced cellular and vascular senescence; this reduction was reversed by metformin. Specifically, metformin enhanced the autophagic flux at the autophagosome-lysosome fusion level, whereas blockade of autophagosome-lysosome fusion inhibited the anti-senescence effects of metformin.

Conclusions: Metformin prevents VSMC and vascular senescence by promoting autolysosome formation.

Keywords: Aging; Autophagic flux; Lysosome; Senescence; VSMC.

Graphical abstract

Fig. 1

Metformin prevents vascular senescence. Mice underwent infusion with saline (control), saline + metformin (150 mg/kg/d), Ang II (400 ng/kg/min) and Ang II (400 ng/kg/min) + metformin (Met, 150 mg/kg/d) with an Alzet osmotic minipump for 14 or 28 days, respectively. A, Upper and middle panels, SA-β-gal staining (positive β-galactosidase staining appears blue) was carried out for assessing senescence of the mouse aorta (14 and 28 days, respectively). A, Lower panel, SA-β-gal staining of aortas from physiological aging mice (n = 3); 10-month-old, 20-month-old, 20-month-old + metformin. B and C, Immunohistochemistry was performed to detect p53 and p21 in aorta samples from 4 groups (n = 3); scale bar = 100 μm. D, E, and F, Immunoblot was performed to assess p53 and p21 protein levels in aorta samples from the four groups (one-way ANOVA, Tukey post hoc; n = 5). α-tubulin was utilized for normalization. Data are mean ± SD. *P < 0.05, ^**P < 0.01, and ^***P < 0.001 versus control group or for indicated comparisons. Met, metformin; Ang II, angiotensin II.

Fig. 2

Metformin alleviates functional and structural alterations during vascular senescence. PWV measurement represents a valid tool for noninvasively quantifying arterial stiffness. A and B, PWV was higher in Ang II-infused mice in comparison with age-matched control animals, while metformin administration decreased PWV in comparison with the Ang II-infusion group (one-way ANOVA, Tukey post hoc; n = 9). Alterations to collagen and elastin contents as well as aortic wall arrangement were assessed by Masson’s Trichrome, Picrosirius Red, and Van Gieson’s staining of aortic tissue ring samples. C, Morphometric assessment of the four mouse groups after Masson’s trichrome staining. F, G, H, I, and J, Medial layer thicknesses, medial cross-sectional areas, lumen diameters, medial thickness/lumen diameter ratios, and collagen levels were calculated (one-way ANOVA, Tukey post hoc; n = 6–7). Scale bar = 200 μm. D and K, Circularly polarized light detection of collagen fibers, with collagen contents obtained within randomly selected fields. Optical density was determined (one-way ANOVA, Tukey post hoc; n = 3). Scale bar = 200 μm. E, Elastin fibers were detected by Van Gieson’s staining; scale bar = 200 μm. Data are mean ± SD. *P < 0.05, ^**P < 0.01, and ^***P < 0.001 versus control group or for indicated comparisons. PWV, pulse wave velocity.

Fig. 3

Metformin treatment delays cellular senescence and alleviates the SASP of VSMCs. Senescent primary VSMCs from the aortas of elderly patients were used to examine metformin’s effect on cellular senescence. A, B and C, The number of SA-β-gal-positive cells as well as p53, p21, and p16 protein levels were decreased by metformin treatment (t-test, n = 3). Ang II-induced premature senescent VMSCs were treated with metformin for further confirm its beneficial effects in cellular senescence. D and E, SA-β-gal staining was carried out for examining senescence in VSMCs (one-way ANOVA, Tukey post hoc; n = 3). F, G, and H, Immunoblot was carried out to determine p53 and p21 protein levels in Ang II-induced senescent VSMCs. GAPDH was utilized for normalization (one-way ANOVA, Tukey post hoc; n = 3–4). I, J, K, and L, SASP markers in Ang II-induced senescent VSMCs, including MMP-2, IL-6, and TGF-β, were further assessed by immunoblot. GAPDH was utilized for MMP-2 and IL-6 level normalization, and α-tubulin was used for TGF-β level normalization (one-way ANOVA, Tukey post hoc; n = 3–4). Data are mean ± SD. *P < 0.05, ^**P < 0.01, and ^***P < 0.001 versus control group or for indicated comparisons.

Fig. 4

Metformin has beneficial effects on function restoration in senescent VSMCs. The migration of Ang II-induced senescent VSMCs was determined by the wound-healing assay. C, Semi-quantitative analysis of wound area healing at a fixed location was performed with Image J. VSMCs underwent incubation with or without Ang II (2 μM) for 24 h with metformin (200 μM) treatment or not (one-way ANOVA, Tukey post hoc; n = 4). B, D, E, and F, The expression levels of PCNA, SM-22α, and calponin in Ang II-induced senescent VSMCs were assessed by immunoblot, using α-tubulin for normalization (one-way ANOVA, Tukey post hoc; n = 3). G-H, Flow cytometry showed G0/G1 cell-cycle arrest was increased significantly in Ang II-treated cells, and this effect was alleviated by metformin treatment, while metformin increased S phase cells that were inhibited by Ang II treatment (one-way ANOVA, Tukey post hoc; n = 3). Data are mean ± SD. *P < 0.05, ^**P < 0.01, and ^***P < 0.001 versus control group or for indicated comparisons.

Fig. 5

Autophagy contributes to metformin’s beneficial effects in senescence. A, Proteomics of aorta specimens from metformin-treated mice revealed 202 upregulated and 252 downregulated proteins, with statistical significance. B and C, GO and COG analyses showed metformin’s targets were markedly enriched in catalytic activity, posttranslational modification, protein turnover, and chaperones. D and E, The autophagosome levels in aorta specimens from the 4 groups were examined by TEM (one-way ANOVA, Tukey post hoc; n = 4). Scale bar = 1.0 µm (upper panel) or 500 nm (lower panel). F and G, LC3, and p62 levels in aorta samples were assessed by immunofluorescence; α-SMA was simultaneously stained to confirm the location of VSMCs. H, I, and J, Immunoblot assessment of LC3 and p62 in aorta specimens, with α-tubulin used for normalization (one-way ANOVA, Tukey post hoc; n = 5). Data are mean ± SD. *P < 0.05, ^**P < 0.01, and ^***P < 0.001 versus control group or for indicated comparisons.

Fig. 6

Metformin promotes autophagic flux at the autophagosome-lysosome fusion level. A, B, and C, The expression levels of LC3 and p62 in Ang II-induced senescent VSMCs were examined by immunoblot. VSMCs underwent incubation in the presence or absence of Ang II (2 μM) for 24 h, with metformin (200 μM) treatment or not. GAPDH was utilized for normalization (one-way ANOVA, Tukey post hoc; n = 3). D and E, To inhibit autophagy, VSMCs were incubated with bafilomycin A1, then treated with metformin and Ang II; LC3 protein levels were assessed by immunoblot, with GAPDH as a loading control (one-way ANOVA, Tukey post hoc; n = 5). F and G, AMPK protein, and phosphorylation levels were evaluated by immunoblot in senescent primary VSMCs treated or not with metformin. H, To evaluate metformin’s effect on autophagic flux, autolysosome staining was performed; scale bar = 100 μm. I, VSMCs was pre-incubated with bafilomycin A1 for two hours, then treated with metformin and Ang II; LC3, p62, p53, and p21 protein levels were evaluated by immunoblot. J-M, LAMP1, CSTL, CSTB protein levels were examined by immunoblot (one-way ANOVA, Tukey post hoc; n = 3–4). Data are mean ± SD. *P < 0.05, ^**P < 0.01, and ^***P < 0.001 versus control group or for indicated comparisons. Baf A1, bafilomycin A1.