Epigenetic modulation of Drp1-mediated mitochondrial fission by inhibition of S-adenosylhomocysteine hydrolase promotes vascular senescence and atherosclerosis

Abstract

Aims: Vascular senescence, which is closely related to epigenetic regulation, is an early pathological condition in cardiovascular diseases including atherosclerosis. Inhibition of S-adenosylhomocysteine hydrolase (SAHH) and the consequent increase of S-adenosylhomocysteine (SAH), a potent inhibitor of DNA methyltransferase, has been associated with an elevated risk of cardiovascular diseases. This study aimed to investigate whether the inhibition of SAHH accelerates vascular senescence and the development of atherosclerosis.

Methods and results: The case-control study related to vascular aging showed that increased levels of plasma SAH were positively associated with the risk of vascular aging, with an odds ratio (OR) of 3.90 (95% CI, 1.17-13.02). Elevated pulse wave velocity, impaired endothelium-dependent relaxation response, and increased senescence-associated β-galactosidase staining were observed in the artery of SAHH^+/- mice at 32 weeks of age. Additionally, elevated expression of p16, p21, and p53, fission morphology of mitochondria, and over-upregulated expression of Drp1 were observed in vascular endothelial cells with SAHH inhibition in vitro and in vivo. Further downregulation of Drp1 using siRNA or its specific inhibitor, mdivi-1, restored the abnormal mitochondrial morphology and rescued the phenotypes of vascular senescence. Furthermore, inhibition of SAHH in APOE^-/- mice promoted vascular senescence and atherosclerosis progression, which was attenuated by mdivi-1 treatment. Mechanistically, hypomethylation over the promoter region of DRP1 and downregulation of DNMT1 were demonstrated with SAHH inhibition in HUVECs.

Conclusions: SAHH inhibition epigenetically upregulates Drp1 expression through repressing DNA methylation in endothelial cells, leading to vascular senescence and atherosclerosis. These results identify SAHH or SAH as a potential therapeutic target for vascular senescence and cardiovascular diseases.

Keywords: Atherosclerosis; DNA methylation; S-adenosylhomocysteine hydrolase; mitochondrial dynamics; vascular senescence.

Graphical abstract

Fig. 1

Plasma SAH levels are associated with vascular aging in population A case‒control study (n = 51/group) related to vascular aging was conducted. Cases were defined as ba-PWV ≥1400 cm/s. (A)–(C) Differences in the levels of plasma SAH, SAM, and SAM/SAH between controls and cases. Plasma levels of SAH and SAM were measured by stable-isotope dilution liquid chromatography–electrospray tandem mass spectrometry. (D)–(F) The correlation between plasma SAH levels and baPWV, cIMT, and FMD in the cases and the controls. ba-PWV, brachial ankle pulse wave velocity; cIMT, carotid intima-media thickness; FMD, flow-mediated dilation. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the Mann-Whitney U test). Correlation was conducted using Spearman correlation analysis.

Fig. 2

SAHH inhibition induces vascular senescence. (A) 8-week-old male SAHH^+/- mice and wild type littermates were fed a chow diet for 24 weeks (n = 6). (B) Aortic pulse wave velocity in abovementioned mouse groups (n = 6). (C) Endothelium-dependent vascular relaxation in response to acetylcholine in the presence or absence of eNOS inhibitor l-NAME, and endothelium-independent vascular relaxation in response to nitroprusside in abovementioned mouse groups (n = 6). (D) SA β-gal staining and CD31 immunohistochemical staining in the aortic arch of abovementioned mouse groups (n = 6). Scale bar = 2 mm, 250 μm, and 250 μm (from the top to bottom). (E) Western blot analysis of p16, p21, and p53 protein expression in the aorta of abovementioned mouse groups (n = 6). (F) HUVECs were treated with ADA (30 μM) or transfected with SAHH siRNA for 48 h. SA β-gal staining of abovementioned HUVECs groups (n = 4). Scale bar = 200 μm. SA β-gal, senescence-associated β-galactosidase; PWV, pulse wave velocity. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the t-test).

Fig. 3

SAHH inhibition promotes mitochondrial fission and elevated mtROS. (A) Morphology of mitochondria detected by transmission electron microscopy, morphological staining of mitochondria using MitoTracker Green dye, and mtROS staininig using MitoSOX dye in HUVECs treated with ADA (30 μM) or transfected with SAHH siRNA for 48 h (n = 3). Scale bar = 2 μm, 25 μm, and 25 μm (from the top to bottom). (B) Quantification of the ratio of mitochondrial width to length according to TEM examination in (A). (C) Quantification of the mitochondrial length according to Mitotracker Green staining in (A). (D) Quantification of mtROS levels according to MitoSOX staining in (A). (E) Morphology of mitochondria in the aortic endothelium detected by transmission electron microscopy and DHE staining of the aorta in wild type and SAHH^+/- mouse groups (n = 6). Scale bar = 2 μm and 250 μm (from the top to bottom). (F) Quantification of the ratio of mitochondrial width to length according to TEM examination in (E). (G) Quantification of DHE fluorescence density according to DHE staining in (E). (H) SA β-gal staining of HUVECs transfected with SAHH-siRNA for 48 h in the presence or absence of mitoTEMPO (50 μM) (n = 4). Scale bar = 200 μm. (I) Male SAHH^+/- mice and their wild type littermates were divided into 4 groups (n = 6): wild type group, wild type + mitoTEMPO (i.p. 0.5 mg/kg every other day), SAHH^+/- group, and SAHH^+/- + mitoTEMPO (i.p. 0.5 mg/kg every other day) group and fed a chow diet from 8 to 32 weeks of age. (J) SA β-gal staining and CD31 immunohistochemical staining of the aortic arch of the abovementioned mouse groups (n = 6). Scale bar = 2 mm, 250 μm, and 250 μm (from the top to bottom). TEM, transmission electron microscopy; SA β-gal, senescence-associated β-galactosidase; mtROS, mitochondria-derived reactive oxygen species. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the t-test or 1-way ANOVA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Enhanced mitochondrial fission and mtROS are mediated by Drp1. HUVECs were treated with ADA (30 μM) or transfected with SAHH siRNA for 48 h. 8-week-old SAHH^+/- mice and wild type littermates were fed a chow diet for 24 weeks (n = 6). (A) Western blot analysis of genes regulating mitochondrial dynamics in abovementioned HUVECs groups (n = 3). (B) Western blot analysis of Drp1 protein expression in the aorta of abovementioned mouse groups (n = 6). HUVECs were transfected with DRP1-siRNA in the presence of ADA or SAHH-siRNA. (C) Western blot analysis of Drp1 protein expression in abovementioned HUVECs groups (n = 3). (D) Morphology of mitochondria detected by transmission electron microscopy, morphological staining of mitochondria using MitoTracker Green dye, and mtROS staininig using MitoSOX dye in abovementioned HUVECs groups (n = 3). Scale bar = 2 μm, 25 μm, and 25 μm (top to bottom). (E) Quantification of the ratio of mitochondrial width to length according to TEM examination in (D). (F) Quantification of the mitochondrial length according to MitoTracker Green staining in (D). (G) Quantification of mtROS levels according to MitoSOX staining in (D). (H) Male SAHH^+/- mice and their wild type littermates were divided into 4 groups (n = 6): wild type group, wild type + mdivi-1 (i.p. 10 mg/kg twice a week) group, SAHH^+/- group, and SAHH^+/− + mdivi-1 (i.p. 10 mg/kg twice a week) group and fed a chow diet from 8 to 32 weeks of age. (I) Morphology of mitochondria in the aortic endothelium detected by transmission electron microscopy and DHE staining of the aorta in abovementioned mouse groups (n = 6). Scale bar = 2 μm and 250 μm (top to bottom). (J) Quantification of the ratio of mitochondrial width to length according to TEM examination in (I). (K) Quantification of DHE fluorescence density according to DHE staining in (I). For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the t-test or 1-way ANOVA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Drp1 inhibition rescues the endothelial senescence and vascular senescence induced by SAHH inhibition. HUVECs were transfected with DRP1-siRNA in the presence of SAHH-siRNA. Male SAHH^+/- mice and their wild type littermates were divided into 4 groups (n = 6): wild type group, wild type + mdivi-1 (i.p. 10 mg/kg twice a week) group, SAHH^+/-group, and SAHH^+/− + mdivi-1 (i.p. 10 mg/kg twice a week) group and fed a chow diet from 8 to 32 weeks of age. (A) SA β-gal staining of HUVECs from the abovementioned 4 groups (n = 4). Scale bar = 200 μm. (B) Western blot analysis of p16, p21, and p53 protein expression in HUVECs from the abovementioned 4 groups (n = 3). (C) Aortic pulse wave velocity of mice from the abovementioned mouse groups (n = 6). (D) Endothelium-dependent vascular relaxation in response to acetylcholine in the presence or absence of eNOS inhibitor l-NAME, and endothelium-independent vascular relaxation in response to nitroprusside in abovementioned mouse groups (n = 6). (E) SA β-gal staining and CD31 immunohistochemical staining of the aortic arch of the abovementioned mouse groups (n = 6). Scale bar = 2 mm, 250 μm, and 250 μm (from the top to bottom). (F)–(G) Western blot analysis of p16, p21, and p53 protein expression in mouse aorta from the abovementioned mouse groups (n = 6). SA β-gal, senescence-associated β-galactosidase; PWV, pulse wave velocity. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the 1-way ANOVA).

Fig. 6

SAHH declines and vascular senescence increases during the progression of atherosclerosis in ApoE^−/− mice. (A) 8-week-old male APOE^−/− mice were divided into 3 groups (n = 6) and fed a western diet for 0, 6 and 12 weeks, respectively. (B) SA β-gal staining and CD31 immunohistochemical staining of the aortic arch and oil red O staining of the aortic sinus of the abovementioned mice (n = 6). Scale bar = 2 mm, 250 μm, 250 μm, and 500 μm (from the top to bottom). (C)–(D) Plasma concentrations of SAH and the SAM/SAH ratios in the abovementioned mice (n = 6). Plasma levels of SAH and SAM were measured by stable-isotope dilution liquid chromatography–electrospray tandem mass spectrometry. (E) Western blot analysis of p16, p21, p53, SAHH, and Drp1 protein expression in the aorta of the abovementioned mice (n = 6). SA β-gal, senescence-associated β-galactosidase. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the 1-way ANOVA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 7

SAHH inhibition accelerates atherosclerosis in a Drp1-dependent pathway. (A) 8-week-old male APOE^−/−SAHH^+/− mice and APOE^−/− littermates were divided into 4 groups (n = 6): APOE^−/−group, APOE^−/− + mdivi-1 (i.p. 10 mg/ml twice a week) group, APOE^−/−SAHH^+/− group, and APOE^−/−SAHH^+/− + mdivi-1 (i.p. 10 mg/ml twice a week) group and fed a western diet from 8 to 16 weeks of age. (B)–(C) Plasma concentrations of SAH and the SAM/SAH ratios in the abovementioned mice (n = 6). Plasma levels of SAH and SAM were measured by stable-isotope dilution liquid chromatography–electrospray tandem mass spectrometry. (D) Representative SA β-gal staining images and oil red O staining images of the same aorta in abovementioned mouse groups (n = 6). (E)–(F) Quantification of the SA β-gal staining and oil red O staining areas in (D). (G) Oil red O staining of aortic sinus sections in the abovementioned mouse groups (n = 6). Scale bar = 500 μm. SA β-gal, senescence-associated β-galactosidase. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the 1-way ANOVA). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 8

SAHH inhibition induces hypomethylation of the DRP promoter and upregulation of Drp1 via inhibition of DNMT1. (A) The methylation level of GpG sites over the promoter region of the DRP1 gene in HUVECs treated with ADA (30 μM) or transfected with SAHH siRNA for 48 h by BA-seq. (B) HUVECs transfected with unmethylated or methylated DRP1 promoter by the CpG methylase Sss1 were treated with ADA (30 μM) or transfected with SAHH siRNA for 48 h. DRP1 promoter activity in abovementioned HUVECs measured using a luciferase reporter kit (n = 3). (C) Western blot analysis of DNMT1, DNMT3A, and DNMT3B in HUVECs treated with ADA (30 μM) or transfected with SAHH siRNA for 48 h (n = 3). (D) Western blot analysis of DNMT1 and Drp1 in HUVECs overexpressing DNMT1 in the presence of ADA and SAHH-siRNA (n = 3). (E) HUVECs transfected with the DRP1 promoter or control plasmid were infected with Adv-CT or Adv-DNMT1 for 24 h and then treated with ADA (30 μM) or SAHH siRNA for another 24 h. DRP1 promoter activity measured by a luciferase reporter kit in abovementioned HUVECs groups (n = 3). TSS, transcription start site; BA-seq, amplicon bisulfite sequencing. For all bar graphs, data are means ± SEM, *p < 0.05 (determined by the t-test or 1-way ANOVA).