Reduction of double-strand DNA break repair exacerbates vascular aging

Abstract

Advanced age is the greatest risk factor for cardiovascular disease (CVD), the leading cause of death. Arterial function is impaired in advanced age which contributes to the development of CVD. One underexplored hypothesis is that DNA damage within arteries leads to this dysfunction, yet evidence demonstrating the incidence and physiological consequences of DNA damage in arteries, and in particular, in the microvasculature, in advanced age is limited. In the present study, we began by assessing the abundance of DNA damage in human and mouse lung microvascular endothelial cells and found that aging increases the percentage of cells with DNA damage. To explore the physiological consequences of increases in arterial DNA damage, we evaluated measures of endothelial function, microvascular and glycocalyx properties, and arterial stiffness in mice that were lacking or heterozygous for the double-strand DNA break repair protein ATM kinase. Surprisingly, in young mice, vascular function remained unchanged which led us to rationalize that perhaps aging is required to accumulate DNA damage. Indeed, in comparison to wild type littermate controls, mice heterozygous for ATM that were aged to ~18 mo (Old ATM +/-) displayed an accelerated vascular aging phenotype characterized by increases in arterial DNA damage, senescence signaling, and impairments in endothelium-dependent dilation due to elevated oxidative stress. Furthermore, old ATM +/- mice had reduced microvascular density and glycocalyx thickness as well as increased arterial stiffness. Collectively, these data demonstrate that DNA damage that accumulates in arteries in advanced age contributes to arterial dysfunction that is known to drive CVD.

Keywords: DNA damage; aging; arterial stiffness; endothelial cell; oxidative stress; senescence; vascular function.

Figure 1

Effect of aging on endothelial cell DNA damage. (A) Representative images of immunofluorescence for the DNA damage marker 53BP1 on human microvascular lung endothelial cells from young (29 ± 1 year old, female) and old (67 ± 1 year old, female) donors. N = 3–4 experimental replicates from 2 young and 2 old donors. (B) Percentage of endothelial cells (EC) containing one or more 53BP1 foci. (C) Number of 53BP1 foci per endothelial cell. Experimental replicates on cells from the same donors are denoted by individual data points with like colors. (D) Representative images of immunofluorescence for the DNA damage marker 53BP1 on mouse microvascular lung endothelial cells from young (2.7 ± 0 mo old, male) and old (27 ± 0 mo old, male) mice. N = 5–8 experimental replicates from 12 young and 12 old mice per group. (E) Percentage of endothelial cells containing one or more 53BP1 foci. (F) Number of 53BP1 foci per endothelial cell. Individual data points with black borders denote females. Individual data points matching group colors denote males. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001. Scale bars are 10 μm.

Figure 2

Impact of reduced double-strand DNA break repair on endothelium-dependent and independent vasodilation. (A) Mesenteric artery dose-response curves to increasing doses of the endothelium-dependent vasodilator acetylcholine in the absence and presence of the nitric oxide synthase inhibitor L-NAME. N = 7–11 per group. (B) Maximal acetylcholine (ACh) vasodilation in mesenteric arteries in the absence and presence of L-NAME. N = 7–11 per group. (C) Mesenteric artery dose-response curves to increasing doses of the endothelium-independent vasodilator sodium nitroprusside. Yg = young. N = 8–10 per group. Individual data points with black borders denote female mice. Individual data points matching group colors denote male mice. ^†p < 0.05 vs. Acetylcholine dose-response curve in the absence of L-NAME from the same group.

Figure 3

Impact of aging and reduced double-strand DNA break repair on arterial DNA damage, senescence, and endothelium-dependent and -independent vasodilation. (A) Representative images of immunofluorescence for the DNA damage marker 53BP1 performed in aortic segments of Old ATM +/+ and Old ATM+/− mice. Green is elastin (ELN) autofluorescence from tunica media. N = 4–6 per group. (B) Percentage of aortic cells containing one or more 53BP1 foci. (C) Aortic mRNA expression of atm and senescence-related genes. N = 5–10 per group. (D) Mesenteric artery dose-response curves to increasing doses of the endothelium-dependent vasodilator acetylcholine in the absence and presence of the nitric oxide synthase inhibitor L-NAME. N = 8–14 per group. (E) Maximal acetylcholine (ACh) vasodilation in mesenteric arteries in the absence and presence of L-NAME. N = 8–14 per group. (F) Mesenteric artery dose-response curves to increasing doses of the endothelium-independent vasodilator sodium nitroprusside. N = 14–16 per group. Individual data points with black borders denote female mice. Individual data points matching group colors denote male mice. ^*p < 0.05 vs. Old ATM +/+ Acetylcholine dose-response curve in the absence of L-NAME. ^†p < 0.05 vs. Acetylcholine dose-response curve in the absence of L-NAME from the same group. Scale bars are 50 μm.

Figure 4

Effect of aging and reduced DNA double-strand break repair on oxidative stress-mediated suppression of vasodilation. (A) mesenteric artery dose-response curves to increasing doses of the endothelium-dependent vasodilator acetylcholine in the absence and presence of the superoxide scavenger TEMPOL. N = 7–11 per group. (B) Maximal acetylcholine (ACh) vasodilation in mesenteric arteries in the absence and presence of TEMPOL. N = 7–11 per group. (C) Mesenteric artery dose-response curves to increasing doses of the endothelium-dependent vasodilator acetylcholine in the absence and presence of the superoxide scavenger TEMPOL. N = 6–11 per group. (D) Maximal acetylcholine (ACh) vasodilation in mesenteric arteries in the absence and presence of TEMPOL. N = 6–11 per group. (E) Carotid artery superoxide production measured via electron paramagnetic resonance spectrometry. N = 6–8 per group. Individual data points with black borders denote female mice. Individual data points matching group colors denote male mice. ^$p < 0.05 vs. Old ATM +/+ Acetylcholine dose-response curve in the absence of TEMPOL, ^†p < 0.05 vs. Acetylcholine dose-response curve in the presence of TEMPOL from the same group, ^#p < 0.05 vs. Old ATM+/+ Acetylcholine dose-response curve in the presence of TEMPOL, ^*p < 0.05, ^***p < 0.001.

Figure 5

Effect of aging and reduced DNA double-strand break repair on micro- and macro-vascular properties. (A) Perfused microvascular density in mesenteric microcirculation. N = 10–15 per group. (B) Perfused microvascular density in mesenteric microcirculation. N = 14–18 per group. (C) Density of perfused capillaries between 5–9 μm in size from mesenteric microcirculation. N = 14–18 per group. (D) Perfused boundary region in capillaries between 5–9 μm in size in mesenteric microcirculation. N = 14–18 per group. (E) Perfused boundary region in all microvessels between 10–25 μm in size in mesenteric microcirculation. N = 14–18 per group. (F) Aortic stiffness assessed via pulse wave velocity. N = 11–17 per group. (G) Positive relation between PWV and percentage of aortic cells containing one or more 53BP1 foci. N = 5 per group. (H) Aortic elastin breaks. N = 5–6 per group. (I) Aortic mRNA expression of extracellular matrix and arterial stiffness-related genes. N = 7–9. Individual data points with black borders denote female mice. Individual data points matching group colors denote male mice. ^*p < 0.05, ^**p < 0.01, ^****p < 0.0001.