Sirt1 extends life span and delays aging in mice through the regulation of Nk2 homeobox 1 in the DMH and LH

Abstract

The mammalian Sir2 ortholog Sirt1 plays an important role in metabolic regulation. However, the role of Sirt1 in the regulation of aging and longevity is still controversial. Here we demonstrate that brain-specific Sirt1-overexpressing (BRASTO) transgenic mice show significant life span extension in both males and females, and aged BRASTO mice exhibit phenotypes consistent with a delay in aging. These phenotypes are mediated by enhanced neural activity specifically in the dorsomedial and lateral hypothalamic nuclei (DMH and LH, respectively), through increased orexin type 2 receptor (Ox2r) expression. We identified Nk2 homeobox 1 (Nkx2-1) as a partner of Sirt1 that upregulates Ox2r transcription and colocalizes with Sirt1 in the DMH and LH. DMH/LH-specific knockdown of Sirt1, Nkx2-1, or Ox2r and DMH-specific Sirt1 overexpression further support the role of Sirt1/Nkx2-1/Ox2r-mediated signaling for longevity-associated phenotypes. Our findings indicate the importance of DMH/LH-predominant Sirt1 activity in the regulation of aging and longevity in mammals.

Figure 1. BRASTO mice show the extension of both median and maximum life span and delay the age-associated mortality rate and the incidence of cancer-dependent death

(A) Kaplan-Meier curves of BRASTO (Tg) and wild-type control (WT) mice in females (top left), males (top right) and combined sex (bottom left). P values shown were calculated by log-rank test. (B) Changes in age-associated mortality rate by plotting linear hazard rate in females, males and combined sex. P values for the differences of fitted slopes (not significant) and y-axis intercepts (shown in each panel) were calculated by ANCOVA between BRASTO and wild-type control mice. (C and D) Accumulating incidents of cancer-dependent deaths (C) and identified causes of death (D) in aged BRASTO and wild-type control mice. P value (top) for the difference of the two genotypes was calculated by log-rank test. P values for the differences of fitted slopes (not significant) and y-axis intercepts (bottom) were calculated by ANCOVA. See also Figure S1.

Figure 2. Aged BRASTO mice display higher physical activity, body temperature, oxygen consumption, and better quality of sleep compared to wild-type control mice

(A) Wheel-running activity of BRASTO (Tg) and wild-type control (WT) mice at 20 months (left) and 4 months (right) of age, respectively. Activity counts per minute at each time point are shown as mean values ± S.E. (***p<0.001 by Wilcoxon matched-pairs singled-ranks test, n=10–12 mice for each genotype). (B and C) Rectal body temperature (B) and oxygen consumption (VO₂) (C) at 20 months of age. Rectal body temperature and VO₂ at each time point are shown as mean values ± S.E. [**p<0.01 by Student’s t-test (B); *p<0.05 by Wilcoxon matched-pairs singled-ranks test (C, 10pm to 5am), n=10–12 mice for each genotype]. (D and E) Levels of delta power in NREM sleep (D) and wake (E) at 20 months of age. Levels of delta power at each time point are shown as mean values ± S.E. (***p<0.001 by Wilcoxon matched-pairs singled-ranks test, n=5–6 mice for each genotype). Shaded area represents the dark time. See also Figure S2.

Figure 3. Aged BRASTO mice show more youthful mitochondrial morphology and function in skeletal muscle and enhanced neural activation specifically in the DMH and LH

(A and B) Morphometric analysis of the soleus muscle from aged BRASTO (Tg) and wild-type control (WT) mice. Representative electron microscope images are shown (A). The number (top) and the size (bottom) of mitochondria were measured (B), and results are shown as mean values ± S.E. (*p<0.05, ***p<0.001, n=3 mice for each genotype, 10–15 sections per mouse). (C and D) mRNA expression levels of genes related to skeletal muscle mitochondrial function [Pgc1 α, Idh3 α and Cytb (C)] and stimulation by the sympathetic nervous system [Adrb2 (D)] in the soleus muscle of young (D, left panel) and aged (C and D, right panel) BRASTO and wild-type control mice at 3pm and 9pm. Each mRNA expression level was normalized to that of WT at 3pm. Results are shown as mean values ± S.E. (*p<0.05, **p<0.01, ***p<0.001, n=4–6 mice for each genotype). (E and F) mRNA expression levels of cFos (E) and Ox2r (F) in the Arc, VMH, DMH, and LH of aged BRASTO and wild-type control mice at 9pm. Results are shown as mean values ± S.E. (*p<0.05, **p<0.01, n=3–4 mice for each genotype). mRNA expression levels in each hypothalamic nucleus were normalized to those in the WT Arc. See also Figure S3.

Figure 4. Sirt1 enhances Ox2r promoter activity by interacting with and deacetylating a homeodomain transcription factor Nkx2-1

(A–C) Luciferase activities driven by the 1-kb and truncated Ox2r promoter fragments (A and B, Ox2r-Luc) or the 260-bp Ox2r promoter or mutant carrying mutations in two Nk2 family binding sites (C, Ox2r-112/251mut) in HEK293 cells co-transfected with a Sirt1 minigene or mutant Sirt1 (H355A) (A and B) and/or Nkx2-1 (pCMV-Nkx2-1) (B and C). Luciferase activities in cells co-transfected with a backbone vector were normalized to 100%. Results are shown as mean values ± S.E. [control vs treatment, *p<0.05, **p<0.01, ***p<0.001 by one-way ANOVA with Tukey-Kramer post hoc test (A–C); pCMV-Nkx2-1 alone vs pCMV-Nkx2-1 with Sirt1 minigene, ¶p<0.05, 1¶¶p<0.01, ¶¶¶p<0.001 by Student’s t-test (B and C); pCMV-Nkx2-1 with Sirt1 minigene vs pCMV-Nkx2-1 with H355A (B) or Ox2r-260bp vs Ox2r-112/251mut (C), §p<0.05, §§p<0.01 by Student’s t-test, n=3–9]. (D) Chromatin-immunoprecipitation for Nkx2-1 in the hypothalamus. Primer set 1 is designed for an unrelated genomic region. Locations of primer sets 2–4 are shown in the upper panel. Nk2 family binding sites are located at −112 and −251 positions with respect to the transcription start site in the Ox2r promoter. (E–G) Co-immunoprecipitation showing the interaction of Sirt1 and Nkx2-1 overexpressed in HEK293 cells (E and F) and in the hypothalamic extracts (G). (H) Deacetylation of Nkx2-1 by Sirt1 in HEK293 cells. (I) Luciferase activity of Ox2r-Luc in HEK293 cells co-transfected with a Sirt1 minigene and/or pCMV-Nkx2-1 or Nkx2-1 mutants carrying different point mutations (K161Q, K182Q, or K215Q). Results are shown as mean values ± S.E. (control vs treatment, *p<0.05, **p<0.01, ***p<0.001 by one-way ANOVA with Tukey-Kramer post hoc test; pCMV-Nkx2-1 alone vs pCMV-Nkx2-1 with Sirt1 minigene or pCMV-Nkx2-1 (K215Q) alone vs pCMV-Nkx2-1 (K215Q) with Sirt1 minigene, ¶p<0.05, ¶¶¶p<0.001 by Student’s t-test, n=6). These three lysines lie within the homeodomain of Nkx2-1 protein as shown in the schematic structure of Nkx2-1. Acetylated lysine residues are shown in red. (J) mRNA expression levels of Nkx2-1 and Ox2r after knockdown of Nkx2-1 in primary hypothalamic neurons. mRNA expression levels were normalized to those in control firefly luciferase (fLuc) knockdown cells for each gene. Results are shown as mean value ± S.E. (control vs treatment, ***p<0.001 by Student’s t-test, n=3). (K) Acetylation levels of Nkx2-1 in hypothalamic extracts at 3pm (light time) and 9pm (dark time). See also Figure S4.

Figure 5. Sirt1, Nkx2-1, and their target gene product Ox2r in the DMH and LH are required to maintain physical activity, body temperature, skeletal muscle gene expression, and quality of sleep during the dark time

(A) Immunofluorescence of Sirt1 (green), Nkx2-1 (red), and merged image (right) in the DMH. White boxes show areas that are magnified at the lower right corners of each panel. (B and C) Wheel-running activity (B) and rectal body temperature (C) in DMH/LH-specific Sirt1, Nkx2-1, or Ox2r knockdown mice compared to control firefly luciferase (fLuc) knockdown mice. Results at each time point are shown as mean values ± S.E. (*p<0.05, **p<0.01, ***p<0.001 by Wilcoxon matched-pairs singled-ranks test (B) or Student’s t-test (C), n=6 mice for each group). (D) mRNA expression levels of skeletal muscle genes Pgc1 α, Idh3 α, Cytb, and Adrb2 in the soleus muscle from each knockdown mouse. Results are shown as mean values ± S.E. (*p<0.05 by one-way ANOVA with Tukey-Kramer post hoc test, n=4 mice for each group). (E) Levels of delta power in NREM sleep. Levels of delta power at each time point are shown as mean values ± S.E. (***p<0.001 by Wilcoxon matched-pairs singled-ranks test, n=3–4 mice for each group). (F–I) Wheel-running activity (F and H) and rectal body temperature (G and I) in DMH/LH-specific Sirt1 knockdown compared to fLuc knockdown in aged BRASTO (Tg) or wild-type control (WT) mice (F and G) or DMH-specific Sirt1 overexpression (OE) compared to control in aged B6 mice (H and I). Results at each time point are shown as mean values ± S.E. [Tg-fLuc vs. WT-fLuc, ##p<0.01; Tg-fLuc vs Tg-Sirt1 (F) or DMH-control vs DMH-Sirt1 OE (H), *p<0.05, ***p<0.001 by Wilcoxon matched-pairs singled-ranks test; *p<0.05 by one-way ANOVA with Tukey-Kramer post hoc test (G) and Student’s t-test (I), n=3–6 mice for each group]. Shaded area represents the dark time. See also Figure S5.

Figure 6. Another line of BRASTO mice that show higher Sirt1 expression through the hypothalamus exhibit the complete lack of life span extension, beneficial physiological phenotypes, and activation of the DMH and LH

(A) Levels of Sirt1 mRNA in the Arc, VMH, DMH and LH of line 1 (left) and line 10 (right) BRASTO mice. Relative expression levels were calculated as the ratio of Sirt1 mRNA in BRASTO mice vs wild-type control mice. (*p<0.05, **p<0.01 by one-way ANOVA with Tukey-Kramer post hoc test, n=3 mice for each genotype). (B) mRNA expression levels of cFos in the Arc, VMH, DMH, and LH at 9pm. Results are shown as mean values ± S.E. (*p<0.05 by Student’s t-test, n=4–6 mice for each genotype). mRNA expression levels in each hypothalamic nucleus were normalized to that in the WT Arc. (C) Kaplan-Meier curves of line 1 BRASTO (Tg) and wild-type control (WT) mice in females (top left), males (top right) and combined sex (bottom left). P values were calculated by log-rank test. (D) Identified causes of death in aged line 1 BRASTO and wild-type control mice. (E and F) Wheel-running activity (E) and rectal body temperature (F) at 20 months of age. Results are shown as mean values ± S.E. (n=6–8 mice for each genotype). Shaded area represents the dark time. (G) mRNA expression levels of Pgcl α, Idh3 α, and Adrb2 in the soleus muscle of aged line 1 BRASTO mice and wild-type control mice at 9pm. Results are shown as mean values ± S.E. (n=3 mice for each genotype). See also Figure S6.

Figure 7. A model for the role of hypothalamic Sirt1 in the regulation of aging and longevity in mammals

In the hypothalamus, specifically in the DMH and LH, Sirt1 upregulates Ox2r expression and neural activation by interacting with and deacetylating Nkx2-1. Enhanced neural activity in the DMH and LH stimulates the sympathetic nervous system and maintains skeletal muscle mitochondrial morphology and function, physical activity, body temperature, and oxygen consumption during aging. This neural activation by Sirt1/Nkx2-1/Ox2r is also critical to preserve quality of sleep during aging. These functional benefits contribute to the maintenance of youthful physiology and promote longevity.