Effects of dietary polyphenols on aspects of cell molecular physiology relevant to metabolic health and ageing

Abstract

Numerous studies indicate a wide range of possible health benefits from consumption of dietary polyphenols, including resveratrol and genistein. These effects include suppression of cancers, antioxidant functions and protection from diet-induced obesity. Previous research in the laboratory using the mouse 3T3-L1 adipocyte model revealed that resveratrol affected the expression of genes involved in lipogenesis. I proposed that genistein, which has some structural similarity to resveratrol and, like resveratrol, can act through the estrogen receptor (ER), would have the same actions. However, I observed that genistein induced a change in the appearance of cells such that they resembled brown, rather than white, adipocytes. This change in appearance included the accumulation of small, multiclocular lipid droplets, typical of brown adipose tissue, which contrasted with the large lipid droplets seen white adipocytes. I thus proposed that genistein can induce the development of beige adipocytes (adipocytes with characteristics of brown adipocytes but derived from the same lineage as white adipocytes). In support of this hypothesis, genistein reduced mRNAs of genes that characterise white adipocytes (Acaca, Fasn, Fabp4, Lipe, Rarres2, Retn) and increased mRNAs of genes expressed specifically in brown adipocytes (Ucp1, Tnfrsf9) and of genes recognised as mediators of white to brown adipocyte interconversion (Sirt1, Cebpb, Ppargc1a) in a dose- and time-dependent manner. Measurement of mitochondrial activity revealed that basal oxygen consumption rate and mitochondrial proton leak were higher in genistein treated cells, consistent with the respiratory characteristics of beige, rather than white, adipocytes. The presence of the ER antagonist fulvestrant reduced none of these effects of genistein on 3T3-L1 cells, indicating that this action of genistein is not through the ER. However, the responses of Ucp1 and Cebpb genes to genistein were significantly attenuated by EX-527, which is a specific and potent Sirt1 inhibitor. A moderate increase (2 fold) in Sirt1 mRNA level was observed in response to treatment of 3T3-L1 cells with genistein. However, there was an increase of approximately 12 fold in Sirt1 protein, which further indicated that increased Sirt1 action is a key mediator in the change I observed. This effect of genistein on Sirt1 expression was not unique to adipocytes; a similar ER-independent effect was measured in MCF-7 breast cancer cells. Again to develop previous work conducted in the laboratory on novel actions of resveratrol, I proposed that genistein would affect histone expression. I observed that genistein repressed transcription from a histone H3.1 (variant H3b) promoter-reporter IV construct but increased the expression of histone H3 and H4 proteins. Genistein had differential actions on specific histone variant mRNAs. Notably histone H3d and H3.3 mRNAs were increased whereas all other variants measured were reduced. The ER antagonists fulvestrant and G15 did not affect any of these responses, indicating that, like the other actions studied in this body of work, these effects of genistein were though pathways independent of ERs. Observations of these ER-independent actions of genistein on histone expression extended also into 3T3-L1 cells. Gene expression through specific alterations in chromatin structure may thus be one of the mechanisms through which genistein has its reported beneficial effects. In further work exploring ER-dependent versus ER-independent actions I determined if genistein had any effects on expression of ERs, since there is evidence that 17β-estradiol has some autoregulatory feedback action via repression of ER expression. Here I made the potentially important observation that genistein has selective actions on ERα versus ERβ, reducing mRNA corresponding to ERα while increasing ERβ mRNA. Again, these responses were not affected by ER antagonists. This action may be of benefit in the prevention or treatment of breast cancer, where ERα activity is generally detrimental while ERβ activity appears to be of benefit. Finally, I developed the previous research on actions of resveratrol in 3T3-L1 cells by determining if the action of resveratrol to reduce mRNAs for the enzymes fatty acid synthase and acetyl Co-A carboxylase, measured previously and that I reconfirmed in the current work, was accompanied by a change in DNA methylation of the corresponding gene promoter regions, which I proposed to be a plausible mechanism of regulation given other reported effects of reseveratrol to alter DNA methylation. However, there was no effect of resveratrol on DNA methylation in the regions measured. Together, the findings reported in this thesis indicate that a diet with a high genistein content may protect against obesity and other features of the metabolic syndrome by encouraging the development of beige, rather than white, adipose tissue. Such actions appear to be through pathways independent of the ER. Alterations in histone expression may be one of the mediating pathways, but at present I have no evidence of any causal link between the observed effects of genistein on histone expression and on adipocyte gene expression profile and phenotype. Despite excluding effects through the ER as a component of these actions of genistein, I obtained preliminary evidence that modifying the expression of ERs may be an action of genistein relevant to reported cancer-protective actions.