Limited role of Sirt1 in cancer protection by dietary restriction

Abstract

Dietary restriction (DR) has multiple beneficial effects, the two most prominently studied being an increased longevity and an increased cancer protection. Mammalian Sirt1 is a protein deacetylase that has been linked to DR. To explore the relation between Sirt1 and DR, we have examined here DR-induced cancer protection in mice overexpressing Sirt1 (2–3-fold) under its own regulatory elements (Sirt1-tg mice). In particular, we have subjected p53-deficient mice, carrying or not the Sirt1-tg allele, to every-other-day fasting (EOD), which is a type of DR that significantly delays cancer onset. As expected, EOD extended the survival of p53-heterozygous (p53^+/−) mice. However, the extension of survival of p53-heterozygous mice by EOD was the same in the presence or absence of the Sirt1-tg allele. these results suggest that Sirt1 has a limited role in mediating cancer protection by DR in mammals.

Introduction

Dietary restriction (DR) increases longevity across different species. In lower organisms, the protein deacetylase Sir2 has been mechanistically linked to the effects of DR. In particular, genetic deletion of Sir2 in yeast¹ and flies² cancels lifespan extension by DR, whereas genetic overexpression of Sir2 in yeast,³ flies² and worms⁴ promotes lifespan extension under normal feeding conditions. In mammals, there is a family of Sir2 paralogs (Sirt1-7), Sirt1 being the closest paralog to the Sir2 protein present in lower organisms. Whole-body knockout mice for Sirt1 respond normally to some key aspects of DR, such as lower serum levels of glucose, triglycerides and Igf1,⁵ however, these mice do not respond behaviorally to DR by increasing their physical activity.^5-7 Moreover, DR protects mice from specific types of brain and kidney damage, but this protective effect of DR is abolished in the absence of Sirt1.^8,9 Together, these observations support the hypothesis that mammalian Sirt1, similarly to Sir2 in lower organisms, might be a key mediator of DR in mammals.

We previously reported that transgenic mice overexpressing Sirt1 (2–3-fold) under the control of its own endogenous promoter (Sirt1-tg) are significantly protected from the metabolic damage induced by high-fat diet.¹⁰ Regarding aging, Sirt1-tg mice show a healthier aging and are protected from several aging-associated pathologies, however, these beneficial effects are not sufficiently potent to extend lifespan.¹¹ It is conceivable that higher levels of Sirt1 overexpression might be required to obtain lifespan extension in mammals or, alternatively, that Sirt1, despite protecting against several aging-associated diseases, might not impact on the key determinants of longevity.

In addition to its effects on longevity, DR has other beneficial effects, the most notable one being its ability to increase cancer protection.¹² Mice deficient for p53 are tumor prone, and this results in a dramatically reduced survival.^13,14 Interestingly, p53-deficient mice, both p53^+/− and p53^−/−, benefit from a significant increase in cancer-free survival under DR.^15,16 To further explore the putative role of Sirt1 in DR, we have crossed Sirt1-tg mice with mice deficient for the tumor suppressor protein p53 (p53^+/−) and analyzed their cancer-free survival.

Results

We have generated cohorts of p53-het (p53^+/−) and p53-het/Sirt1-tg mice. The survival curves of both cohorts were indistinguishable under ad libitum (AL) conditions (Fig. 1), which suggests that Sirt1 overexpression does not result in cancer protection in a p53-heterozygous background. To test the role of Sirt1 in DR-induced cancer protection, we subjected both mouse lines to every-other-day fasting (EOD) starting at 1 month of age. EOD has been widely used as a model for DR and has been shown to effectively increase lifespan in mice.¹⁷ In particular, EOD increases Sirt1 expression¹⁸ and delays tumor onset in p53-heterozygous mice.¹⁵ Indeed, we observed that p53-heterozygous mice under EOD had an increased survival compared to the AL group (Fig. 1). However, EOD was equally effective in delaying cancer in p53-het and in p53-het/Sirt1-tg mice (Fig. 1). We conclude that Sirt1 overexpression (2–3-fold) does not participate in DR-elicited cancer protection.

Figure 1.

Lack of phenotypic interaction between Sirt1 overexpression and dietary restriction. Survival of the indicated cohorts of mice (p53-het, black lines; p53-het/Sirt1-tg red lines) under ad libitum feeding (AL, dashed lines) and under every-other-day fasting (EOD, solid lines). Cohort sizes were: p53-het AL, n=26; p53-het/Sirt-tg AL, n=23; p53-het EOD, n=17; p53-het/Sirt1-tg EOD, n=9. Statistical significance was assessed using the logrank test.

Discussion

We report here two observations regarding the effect of genetic overexpression of Sirt1 (2–3-fold) in mice. Under AL conditions, we observe that the Sirt1-tg allele has no effect on the survival of p53-heterozygous mice. Previous investigators have reported that Sirt1 overexpression in lymphocytes protects irradiated p53-heterozygous mice from cancer.¹⁹ It can be argued that this protection is due to the effect of Sirt1 on genomic stability following DNA damage,^19,20 rather than to the effect of Sirt1 on tumorigenesis driven by p53 deficiency. In a different study, Sirt1-deficiency was shown to accelerate p53-heterozygous tumorigenesis.²⁰ It is conceivable that wild-type levels of Sirt1 are maximally protective for this type of tumorigenesis and, therefore, further increases in Sirt1 levels do not translate into additional protection, while reduced Sirt1 levels favor cancer.

The main goal of our study was to evaluate the phenotypic interaction between DR and genetic overexpression of Sirt1. However, Sirt1 overexpression did not increase the survival of p53-heterozygous mice when subjected to EOD, and p53-heterozygous mice benefited equally from EOD regardless of the presence or absence of the Sirt1-tg allele. It could be argued that the level of Sirt1 overexpression in our Sirt1-tg mice is not sufficiently potent to synergize with DR. This is certainly possible, although it must be noted that our Sirt1-tg mice present a number of robust phenotypes, most notably including a potent protection from liver steatosis associated to high-fat diet¹⁰ and a dramatic protection from liver damage produced by diethylnitrosamine.¹¹ Therefore, we conclude that Sirt1 plays no role, or a limited role, in DR-induced cancer protection.

Materials and Methods

Animal experimentation

The mouse line Sirt1-tg was previously described,^10,11 and the p53-heterozygous mice have been also described.¹⁴ Mice were housed at the specific pathogen free (SPF) barrier area of the Spanish National Cancer Research Center (CNIO), Madrid and were treated in accordance with the Spanish Laws and the Guidelines for Humane Endpoints for Animals Used in Biomedical Research. For every-other-day fasting (EOD), mice had ad libitum access to food during one day (from early morning to early morning of the following day), and the next day the mice had no food at all (from early morning to early morning of the following day). Cages were also changed together with the removal of the food, in order to avoid any possible food remaining in the bedding. This routine was maintained since mice were 1 month old until death. Mice were observed on a daily basis and sacrificed when they showed signs of morbidity or overt tumors.

Statistical analyses

Kaplan-Meier curves were compared using the logrank test.