Ovariectomy uncouples lifespan from metabolic health and reveals a sex-hormone-dependent role of hepatic mTORC2 in aging

Abstract

Inhibition of mTOR (mechanistic Target Of Rapamycin) signaling by rapamycin promotes healthspan and longevity more strongly in females than males, perhaps because inhibition of hepatic mTORC2 (mTOR Complex 2) specifically reduces the lifespan of males. Here, we demonstrate using gonadectomy that the sex-specific impact of reduced hepatic mTORC2 is not reversed by depletion of sex hormones. Intriguingly, we find that ovariectomy uncouples lifespan from metabolic health, with ovariectomized females having improved survival despite paradoxically having increased adiposity and decreased control of blood glucose levels. Further, ovariectomy unexpectedly promotes midlife survival of female mice lacking hepatic mTORC2, significantly increasing the survival of those mice that do not develop cancer. In addition to identifying a sex hormone-dependent role for hepatic mTORC2 in female longevity, our results demonstrate that metabolic health is not inextricably linked to lifespan in mammals, and highlight the importance of evaluating healthspan in mammalian longevity studies.

Keywords: aging; genetics; genomics; healthspan; human biology; mTOR; mTORC2; medicine; mouse; ovariectomy; sex.

Figure 1.. Deletion of hepatic Rictor impairs male survival independently of sex hormones.

(A) Kaplan-Meier plot of the survival of male mice in which cancer was observed during gross necropsy (N = 38 male mice; WT Sham 5, L-RKO Sham 9, WT CAST 10, L-RKO CAST 14; Supplementary file 1). (B) Kaplan-Meier plot of the survival of male mice in which cancer was not observed during gross necropsy (N = 32 male mice; WT Sham 7, L-RKO Sham 11, WT CAST 11, L-RKO CAST 3; Supplementary file 1). (C) Kaplan-Meier plot of the survival of male mice lacking hepatic Rictor (L-RKO) and their wild-type (WT) littermates. All mice were subjected to gonadectomy (CAST) or Sham surgery at 3 weeks of age (N = 105 male mice; WT Sham 23, L-RKO Sham 27, WT CAST 29, L-RKO CAST 26; Supplementary file 1). (A–C) The overall effect of genotype (RKO), gonadectomy (CAST), and the interaction was determined using a Cox-proportional hazards test (HR, hazard ratio). The two-sided log-rank sum p-value was then calculated comparing individual curves for effects identified as significant in the regression analysis, and corrected for multiple comparisons (Holm-Sidak).

Figure 1—figure supplement 1.. Frequency of cancer observed at necropsy in male mice.

Frequency at which cancer was observed (bottom) or not observed (top) in male mice. N = 70 male mice; p-value was determined by chi-square analysis.

Figure 2.. Ovariectomy protects female mice lacking hepatic Rictor during midlife.

(A) Kaplan-Meier plot of the survival of female mice in which cancer was observed during gross necropsy (N = 37 female mice; WT Sham 11, L-RKO Sham 9, WT OVX 10, L-RKO OVX 7; Supplementary file 1). (B) Kaplan-Meier plot of the survival of female mice in which cancer was not observed during gross necropsy (N = 36 female mice; WT Sham 11, L-RKO Sham 8, WT OVX 7, L-RKO OVX 10; Supplementary file 1). (C) Kaplan-Meier plot of the survival of female mice lacking hepatic Rictor (L-RKO) and their wild-type (WT) littermates. All mice were subjected to gonadectomy (OVX) or Sham surgery at 3 weeks of age (N = 115 female mice; WT Sham 31, L-RKO Sham 29, WT OVX 27, L-RKO OVX 28; Supplementary file 1). (A–C) The overall effect of genotype (RKO), gonadectomy (OVX), and the interaction was determined using a Cox-proportional hazards test (HR, hazard ratio). The two-sided log-rank sum p-value was then calculated comparing individual curves for effects identified as significant in the regression analysis, and corrected for multiple comparisons (Holm-Sidak).

Figure 2—figure supplement 1.. Frequency of cancer observed at necropsy in female mice.

Frequency at which cancer was observed (bottom) or not observed (top) in female mice. N = 73 female mice; p-value was determined by chi-square analysis.

Figure 3.. Gonadectomy affects weight and body composition, and rescues a male-specific effect of hepatic Rictor deletion on fat mass.

(A, B) The weight of A) male and (B) female mice lacking hepatic Rictor (L-RKO) and their wild-type (WT) littermates was tracked starting at 3 months of age (n varies by month; maximum female N = 22–26 mice/group, maximum male N = 20–23 mice/group). (C–F) The lean mass (C, D) and fat mass (E, F) of mice was determined every 4 months starting at 8 months of age (n varies by month; max female N = 9–17/group; max male N = 12–17/group). p-values for the overall effect of genotype (GT) or surgical treatment (Surgery) represent the p-value from pairwise two-way ANOVA testing. Error bars represent SEM.

Figure 4.. Gonadectomy alters energy balance and fuel source utilization.

(A–D) Metabolic chambers were utilized to determine (A) food consumption, (B) Respiratory exchange ratio (RER), (C) Spontaneous activity, and (D) Energy expenditure in (left) female and (right) male mice lacking hepatic Rictor (L-RKO) and their wild-type (WT) littermates of the indicated surgical treatments at 12 months of age. The overall effect of genotype (RKO), gonadectomy (OVX or CAST), and the interaction represent the p-value from a two-way ANOVA conducted separately for the light and dark cycles; *=p < 0.05 from a Holm-Sidak post-test examining the effect of parameters identified as significant in the two-way ANOVA (n = Females, 10 mice/group; Males, WT Sham = 11 mice; WT CAST = 10 mice, L-RKO Sham = 9 mice, L-RKO CAST = 8 mice). Error bars represent SEM.

Figure 5.. Independent effects of hepatic Rictor deletion and gonadectomy on glucose tolerance and insulin resistance.

(A–D) The glucose tolerance of (A, B) male and (C, D) female mice lacking hepatic Rictor (L-RKO) and their wild-type (WT) littermates, of the indicated ages and surgical treatments, was determined following an overnight fast. Area under the curve: the overall effect of genotype (RKO), gonadectomy (OVX or CAST), and the interaction represent the p-value from a two-way ANOVA; *=p < 0.05 from a Holm-Sidak post-test examining the effect of parameters identified as significant in the two-way ANOVA (n = Males: A) 9–12 mice/group, 3–5 months of age; B) 7–12 mice/group, 10–12 months of age. Females: C) 7–11 mice/group, 3–5 months of age; D) 7–11 mice/group, 10–12 months of age). (E–F) fasting blood glucose and insulin values were used to calculate HOMA2-IR in E) male and F) female 14 month old mice (n = 4–5 mice/group; the overall effect of genotype (RKO), gonadectomy (OVX or CAST), and the interaction represent the p-value from a two-way ANOVA; *=p < 0.05 from a Holm-Sidak post-test examining the effect of parameters identified as significant in the two-way ANOVA). Error bars represent SEM.

Figure 5—figure supplement 1.. Effect of hepatic Rictor loss and gonadectomy on insulin tolerance.

(A–D) The insulin tolerance of (A, B) male and (C, D) female mice lacking hepatic Rictor (L-RKO) and their wild-type (WT) littermates, of the indicated ages and surgical treatments was determined following an overnight fast. Area under the curve: the overall effect of genotype (RKO), gonadectomy (OVX or CAST), and the interaction represent the p-value from a two-way ANOVA; *=p < 0.05 from a Holm-Sidak post-test examining the effect of parameters identified as significant in the two-way ANOVA. (n = Male: A) 9–12 mice/group, 4–5 months of age; (B) 8–9 mice/group, 12–14 months of age. Females: (C) 8–11 mice/group, 3.5–6 months of age; (D) 7–10 mice/group, 12–14 months of age). Error bars represent SEM.

Figure 5—figure supplement 2.. Effect of hepatic Rictor loss and gonadectomy on glucose stimulated insulin secretion.

Glucose stimulated insulin levels were measured 15 min after I.P. administration of 1 g/kg glucose (n = 4–5 mice/group; the overall effect of genotype (RKO), gonadectomy (OVX or CAST), and the interaction represent the p-value from a two-way ANOVA). Error bars represent SEM.