SirT1 directly deacetylates lysine-382 of p53, thereby regulating its function14
, and p53 in turn regulates cell turnover, apoptosis, and senescence15,16
. Our novel data showing that HFHS diet administered to normal mice increases endothelial cell acetylation of p53 and caspase-3 cleavage in the atherosclerotic lesion-prone aortic arch is consistent with acetylation of the SirT1-dependent site on p53 contributing to increased turnover of the endothelium in that site. Although systemic risk factors like hyperlipidemia and hyperglycemia affect their extent, the location of atherosclerotic lesions is dictated primarily by local shear forces exerted by turbulent blood flow patterns 17,3,18
. The inner curvature of the aortic arch is one such site where disturbed flow has been implicated in increasing inflammatory and antioxidant genes in both pigs 17,3,4,18
and mice 4
fed a normal diet. In the mouse fed normal diet, the inner curvature of the aortic arch also has been noted to have increased expression of NFκB p655
protein and adhesion molecule mRNA 4
, as well as decreased expression of endothelial nitric oxide synthase mRNA and protein 19
. In pigs fed a normal diet, the lesion-prone area has been demonstrated to be a site of increased permeability1,18
and increased endothelial cell turnover2
. Consistent with increased cell turnover, the endothelium of the lesion-prone pig aortic arch has an almost 3-fold higher expression of caspase-3 mRNA compared to low prone areas3
Our observations in normal mice fed HFHS diet imply that increased acetylation of the SirT1-dependent site on p53 activity accelerates these changes in endothelial cell phenotype in lesion-prone areas of the aorta, and also suggest that similar mechanisms may contribute to the increased atherogenesis observed in these aortic sites in hyperlipidemic LDLr-/- mice fed the same diet. The fact that hydrogen peroxide and TNFα caused similar changes in HAEC in culture not only indicates that these atherogenic mediators increase acetylation of the SirT1-dependent site on p53 and caspase-3 cleavage, but is also consistent with their potential role in causing the same changes in vivo in lesion-prone aortic endothelium of normal mice fed HFHS diet.
Our previous findings indicate that the polyphenol, S17834, activates AMPK in hepatocytes, accounting for a decrease in lipid synthesis 11,10
. AMPK phosphorylation was decreased in the liver of type 1 diabetic mice, and treatment with S17834 increased its phosphorylation, explaining the decrease in hepatic and serum lipids10
. We have found that AMPK phosphorylation is also decreased in the livers of the type-2 diabetic LDLr -/- mice fed HFHS diet studied here, and the decrease is prevented by S17834 treatment (data not shown). This suggests that a decrease in plasma cholesterol, which has been shown previously to be primarily in the LDL fraction in LDL receptor -/- mice fed HFHS diet12
, is caused by hepatic effects of the polyphenol and explains the decrease in atherosclerotic lesions. The large changes in plasma lipids caused by HFHS diet and S17834 treatment in the LDLr-/- mouse are likely due to the lack of hepatic reuptake of LDL via the LDL receptor, and therefore reflect changes in hepatic lipid synthesis12
Because of these systemic effects, we sought to determine what local effects the polyphenol treatment may have on atherogenic events in the aortic endothelium. This was the principal reason why we studied wild type mice in which changes in blood lipids were minimal and where there were no inflammatory cells observed in the aortic arch of animals fed HFHS diet. The polyphenol did not decrease serum lipids in normal mice fed HFHS diet, excluding an effect on serum lipid levels as a cause for the improvements observed. We focused on p53 acetylation and caspase-3 cleavage, because of the recognized increase in endothelial cell turnover of endothelial cells in the lesion prone aortic arch in atherogenesis20,21
, and showed that both apoptotic signaling components are detectable in endothelial cells of both LDLr-/- and C57BL6 mice fed HFHS diet. We found that S17834 prevented the effects of the HFHS diet, decreasing p53 acetylation at lysine-382 and caspase-3 cleavage in the lesion prone arch of the C57BL6 mice. Because of the confounding influence of changes in plasma lipids, we did not perform a detailed analysis of changes in apoptotic cell signaling in the LDLr-/- mice, but we speculate that similar directional changes accompanied the changes in lesions. Increased apoptosis of fibroblasts, smooth muscle cells, or macrophages can conceivably be pro- or anti-atherosclerotic, and apoptotic cell clearance may regulate atherosclerotic lesion size21
. It is more accepted that increased endothelial cell apoptosis and turnover in the lesion prone aortic arch is an early precursor of atherosclerosis21,20
. This makes more relevant our findings that the polyphenol decreases the degree of apoptotic signaling in endothelial cells in the lesion prone aortic arch of normal mice fed HFHS diet.
Unlike with short-term treatment of HAEC, the polyphenol increased SirT1 protein expression in aortic endothelial cells in chronically treated mice, further supporting that the decreased p53 acetylation and cleaved caspase-3 are accounted for by an increase in SirT1 deacetylase activity in the endothelium. We have not excluded the possibility that activity of histone acetylases or acetyltransferases might affect p53 acetylation in HAEC or the mouse aortic arch which could account for our findings, although in the case of the aortic arch in vivo
, the increased protein expression is consistent with an increase in SirT1 deacetylase activity. While this work was in progress, it was reported that endothelial specific overexpression of SirT1 inhibits development of atherosclerosis in apolipoprotein E-deficient mice7
, providing direct evidence that increased endothelial SirT1 expression can attenuate atherogenesis. It is also possible that sumoylation 22
, phosphorylation 23
, or S
post-translational modifications could modify SirT1 deacetylase activity independently of changes in expression. In addition, the effects of SirT1 may be exerted via stimulation of nitric oxide synthase activity 25
which might also affect endothelial cell turnover.
As with any pharmacological study, the dose used in vivo and the concentrations used in vitro are potential limitations. Plasma levels of S17834, which is largely metabolized in the liver, have not been measured in mice administered the dose given in this study, but we were able to show direct affects of the compound on p53 acetylation, caspase-3 cleavage and apoptosis in submicromolar concentrations applied to HAEC in culture suggesting that the in vivo effects were exerted locally. In addition, studies of polyphenols in particular are complicated because they affect many metabolic regulators including PPAR's, AMP kinase, and mitochondrial electron transport, making it difficult to identify specific effects. Whether polyphenols activate SirT1 in vivo is an intense area of current interest. This underscores the importance of our observation that acetylation of the SirT1-dependent lysine on p53 is increased by HFHS diet and that this can be explained by a local effect of S17834 to increase SirT1 expression at atherosclerotic lesion prone sites.