An investigation of ageing-related genomic effects of resveratrol

Abstract

Dietary restriction (DR) increases lifespan robustly in diverse species. Effects of the dietary polyphenol resveratrol consistent with delayed ageing and/or extension of lifespan have been reported. The involvement in the longevity response to DR of the protein Sirt1, which may be activated by resveratrol and deacetylates a range of cellular substrates that includes histone proteins, identifies epigenetic processes as a pathway that may mediate effects of both DR and dietary resveratrol in delaying ageing and/or extending lifespan. Based on a preliminary observation, the hypothesis underlying the study is that some of the beneficial effects of resveratrol on lifespan/aging are mediated through effects on histone expression that oppose changes observed in ageing. A secondary hypothesis, based on a degree of structural similarity between resveratrol and 17β-oestradiol, was that epigenetic effects of resveratrol are mediated through the estrogen receptor (ER). The effect of resveratrol on histone protein expression was investigated in human intestinal Caco-2 cells and human MCF-7 breast cancer cells. Histone H2a, H2b, H3 and H4 expression was decreased in response to resveratrol treatment in both cell lines. In support of our hypothesis that resveratrol affects ageing through reversing ageing-associated changes in histone proteins, higher levels of H2A, H2B, and H4 expression were detected by western blotting in the small intestine of old (38 months) mice than in younger (12 months) mice. To investigate possible consequences of effects of resveratrol, including effects resulting from altered histone expression, we studied the effect of resveratrol on global gene expression in Caco-2 and MCF-7 cells to address several objectives including: (1) investigating if resveratrol has an effect similar to that of DR at the level of gene expression; (2) identifying if genes or pathways affected by resveratrol were also affected by manipulation of the expression level of Sirt1. For both cell types, the number of genes in the intersection between those affected by resveratrol and a compiled list of genes reported in other studies to respond to DR was greater than expected by chance, supporting the view that responses to resveratrol and to dietary restriction have some commonality and that resveratrol may mimic some effects of dietary restriction. We also found that there was very little overlap between genes affected by resveratrol treatment and by knockdown of Sirt1 expression in Caco-2 cells, which adds to accumulating evidence that resveratrol does not act through effects on Sirt1. To investigate if effects of resveratrol - in particular the reduction in histone protein expression - are mediated through the estrogen receptor (ER), Caco-2 and MCF-7 cells were treated with resveratrol in the presence or absence of the ER antagonist fulvestrant, then total cell lysate was analysed by western blotting. The reduction in histone protein (H2a, H2b, H3 and H4) expression was attenuated by fulvestrant, indicating that resveratrol reduced histone expression via an ER-dependent mechanism. For further investigation of effects of resveratrol on histone expression, Caco-2 cells were transfected with a promoter reporter construct comprising the histone H3 promoter upstream of the β- galactosidase reporter gene, and the effect on reporter gene expression of treatment with resveratrol in the presence and absence of the fulvestrant was measured. Resveratrol reduced reporter gene expression and this effect was attenuated by fulvestrant, demonstrating that resveratrol acts to reduce histone H3 expression at the level of transcription through an ER-mediated mechanism. To investigate if the response to resveratrol treatment is through interaction with estrogen response elements (EREs) in the histone H3 promoter we replaced three potential EREs within the histone H3 promoter region included in the promoter-reporter construct with random sequence. Caco-2 cells were then transfected with either original or mutated promoter-reporter construct and treated with resveratrol or the endogenous ER ligand 17-β estradiol in the presence and absence of fulvestrant. Resveratrol and 17-β estradiol both reduced reporter gene expression from both promoter reporter constructs and in all cases responses were attenuated by fulvestrant, indicating that effects of neither compound, although mediated through the ER, are on the specific sequences region we identified and replaced. In conclusion, these data indicate that resveratrol reduces histone expression in both intestinal and breast cancer cells through an ER-mediated mechanism acting at the level of transcription and that this effect may oppose an accumulation of histone proteins (observed in mouse small intestine) that accompanies ageing. With respect to effects on gene expression, resveratrol was found to mimic some effects of dietary restriction but appeared to act through a mechanism independent of Sirt1.