Short-term calorie restriction in male mice feminizes gene expression and alters key regulators of conserved aging regulatory pathways

Abstract

Background: Calorie restriction (CR) is the only intervention known to extend lifespan in a wide range of organisms, including mammals. However, the mechanisms by which it regulates mammalian aging remain largely unknown, and the involvement of the TOR and sirtuin pathways (which regulate aging in simpler organisms) remain controversial. Additionally, females of most mammals appear to live longer than males within species; and, although it remains unclear whether this holds true for mice, the relationship between sex-biased and CR-induced gene expression remains largely unexplored.

Methodology/principal findings: We generated microarray gene expression data from livers of male mice fed high calorie or CR diets, and we find that CR significantly changes the expression of over 3,000 genes, many between 10- and 50-fold. We compare our data to the GenAge database of known aging-related genes and to prior microarray expression data of genes expressed differently between male and female mice. CR generally feminizes gene expression and many of the most significantly changed individual genes are involved in aging, hormone signaling, and p53-associated regulation of the cell cycle and apoptosis. Among the genes showing the largest and most statistically significant CR-induced expression differences are Ddit4, a key regulator of the TOR pathway, and Nnmt, a regulator of lifespan linked to the sirtuin pathway. Using western analysis we confirmed post-translational inhibition of the TOR pathway.

Conclusions: Our data show that CR induces widespread gene expression changes and acts through highly evolutionarily conserved pathways, from microorganisms to mammals, and that its life-extension effects might arise partly from a shift toward a gene expression profile more typical of females.

Figure 1. Plot of significance versus fold change of all significant genes.

Plot of significance (absolute value of d, |d|) versus log2 fold change, for all 3,855 genes called significant at q<0.1. Ddit4, the gene with the highest significance score, is shown at upper right.

Figure 2. CR-induced phosphorylation changes in Eif4ebp1.

Western blots of unphosphorylated Eif4ebp1 (4EBP1) and Thr69 phosphorylated Eif4ebp1 (p4EBP1) for calorie restricted (CR) and high-calorie-fed (HIGHCAL) mice. Labels at top indicate individual mouse sample codes. CR, Calorie Restriction; FHC, Fixed High Calorie feeding; TAL, True Ad Libitum feeding. TAL1, TAL3, and TAL4 are from livers used to make the RNA pool TAL-P. CR1, CR7, and CR8 are from livers used to make the RNA pool CR-P1. CR4, CR5, and CR6 are from livers used to make the RNA pool CR-P2. FHC3 and FHC3B are separate samplings of the same liver.

Figure 3. GO categories with largest relative fractions of upregulated or downregulated genes.

Each stacked bar graph displays the relative upregulated or downregulated fractions of the total number of genes (n = 100%) significantly altered by CR within a given GO category returned by FuncAssociate (adjusted p<0.05). The upregulated fraction is shown in red and the downregulated fraction is shown in blue.

Figure 4. CR feminizes overall gene expression.

Fractional distributions of all genes changed by CR and which display sexually dimorphic expression in the data of Yang et al. Percent numbers are percent of all genes displaying both specific directional changes from CR and sexually dimorphic expression and corresponding numbers of genes in each of these four classes are given in parentheses. Deviation from expected frequencies was determined by a chi-square test for upregulated genes and separately for downregulated genes, at each q-value cutoff.