Fig 1.
Effect of the HG-induced “memory” on cell growth and aging in HUVECs.
HUVECs were cultured in normal glucose media (D-glucose, 5 mM) for 6 days (NG), high glucose media (30 mM) for 6 days (HG), 3 days of normal glucose supplemented with D-mannitol (25 mM) followed by another 3 days of normal glucose media (MN), or 3 days of high glucose followed by another 3 days of normal glucose media (HN). (A) Cells were fixed and stained for SA β-Gal activity. Histogram represent the percentage of SA β-Gal positive cells (n = 5 image per group). (B) Cell count per microscopic field at different time points during cell subculture. (n = 3 image per group) (C) Relative ROS production. (D-E) Immunoblotting and quantification of SIRT1 expression. Values were normalized to GAPDH. (F) SIRT1 deacetylase activtity. (G-K) Immunoblotting and quantification of p300, SIRT1, p21, and Ac-p53 protein expression. For p300, SIRT1, and p21, values were normalized to β-actin; for Ac-p53, normalized to total p53. Value was normalized to total p53. *P < 0.05.
Fig 2.
Overexpression of SIRT1 or suppression of p300 inhibited transient HG-induced senescent “memory” in HUVECs.
Cells were cultured in HG for 3 days followed by 3 days of NG media. SIRT1 plasmid (pSIRT1) or p300 siRNA (si-p300) was transfected into HUVECs twice, 12 hours prior to HG and NG incubation, respectively. After 6 days of culture, cells were harvested for further analysis. Immunoblotting and quantification of p300, SIRT1, p21, Ac-p53, and p53 protein expression after pSIRT1 transfection (A-F) or after 300 siRNA transfection (G-L). For p300, SIRT1, and p21, values were normalized to β-actin; for Ac-p53, normalized to total p53. SIRT1 activity, SA β-Gal positivity, and relative ROS production in HUVECs transfected with pSIRT1 (M-O) or p300 siRNA (P-R). *P < 0.05.
Fig 3.
Effect of resveratrol and/or metformin on SIRT1-mediated signaling and senescent “memory” in HUVECs.
Immunoblotting and quantification of SIRT1 protein expression after (A-B) resveratrol treatment (RSV; 0, 1.25, 2.5, 5, or 10 μM) or (C-D) metformin treatment (MET; 0, 1, 10, 50, 100, 250 μM). Values were normalized to β-actin. (E-J) Immunoblotting and quantification of p300, SIRT1, p21, Ac-p53, and p53 protein expression after RSV (5 μM) or MET (50 μM) treatment. For p300, SIRT1, and p21, values were normalized to β-actin; for Ac-p53, normalized to total p53. (K) The deacetylase activity of SIRT1 in HUVECs treated with or without RSV (5 μM) or MET (50 μM). (L) The percentage of SA β-Gal positive cells in HUVECs treated with or without RSV (5 μM) or MET (50 μM). *P < 0.05.
Fig 4.
SIRT1 knockdown abolished the protective effects of resveratrol and metformin against the occurrence of senescent “memory”.
HUVECs were transfected with SIRT1 siRNA (si-SIRT1) and incubated in HG for 3days and then NG for 3 days. Resveratrol (RSV) or metformin (MET) was added into culture media during the entire 6 days of subculture. Immunoblotting and quantification of p300, SIRT1, p21, Ac-p53, and p53 protein expression with RSV (A-F) or MET treatment (G-L). For p300, SIRT1, and p21, values were normalized to β-actin; for Ac-p53, normalized to total p53. The deacetylase activity of SIRT1 with RSV (M) or MET (N). The percentage of SA β-Gal positive cells with RSV (O) or MET (P). *P < 0.05.
Fig 5.
AMPK knock down abolished resveratrol- and metformin-mdiated SIRT1 activation.
HUVECs were transfected with AMPK siRNA (si-AMPK) and incubated in HG for 3days and then NG for 3 days. Resveratrol (RSV) or metformin (MET) was added into culture media during the entire 6 days of subculture. Immunoblotting and quantification of p300, SIRT1, p21, Ac-p53, and p53 protein expression (A-G). For p300, SIRT1, and p21, values were normalized to β-actin; for Ac-p53, normalized to total p53. (H) Relative ROS production. (I) Deacetylase activity of SIRT1. (J) SA β-Gal positive positivity. *P < 0.05.
Fig 6.
Effect of treatment duration and time points on senescent “memory”.
HUVECs in HN group were treated with RSV or MET during the first 3 days of HG incubation (RSV1-3 or MET1-3), the last 3 days of NG incubation (RSV4-6 or MET4-6), or during 3 days of HG followed by 3 days of NG incubation (RSV1-6 or MET1-6). HUVECs in HN group without treatment of RSV or MET were used as control. (A-C) Immunoblotting and quantification of SIRT1 and SMP-30 protein expression in HUVECs treated with or without RSV. Values were normalized to β-actin. (D) The percentage of SA β-Gal positive cells in HUVECs treated with or without RSV. (E-G) Immunoblotting and quantification of SIRT1 and SMP-30 protein expression in HUVECs treated with or without MET. Values were normalized to β-actin. (H) The percentage of SA β-Gal positive cells in HUVECs treated with or without MET. *P < 0.05 compared to HN without RSV or MET treatment.
Fig 7.
Schematic overview summarizing the molecular mechanisms of resveratrol (RSV) and metformin (MET) against the “metabolic memory” of endothelial senescence in HUVECs.
SIRT1 (a histone deacetylase, HDAC) could directly promote the deacetylation of acetylated (Ac)-p53 or indirectly suppress the acetylation of nascent p53 through inhibition of p300 (a histone acetyltransferase, HAT) expression, and thereby modulate Ac-p53/p21-mediated endothelial senescence. Incubating cells in High glucose (HG) for 3 days followed by normal glucose (NG) for 3 days (a cellular model of transient hyperglycaemia) enhanced intracellular ROS production, which in turn inhibited the expression and deacetylase activity of SIRT1 and simultaneously increased the protein level of p300. Accordingly, the transcriptional activity of p53 was increased due to sustained acetylation at lysine 382 (K382), and the expression of p21, a downstream gene of p53, was activated, leading to persistent endothelial senescence (a cellular model of “metabolic memory”). Furthermore, resveratrol (RSV) and metformin (MET) could prevent the transient hyperglycaemia-induced senescent “memory” via AMPK-dependent ROS reduction and SIRT1 activation.