Comparison of rapamycin schedules in mice on high-fat diet

Abstract

At a wide range of doses, rapamycin extends life span in mice. It was shown that intraperitoneal injections (i.p.) of rapamycin prevent weight gain in mice on high-fat diet (HFD). We further investigated the effect of rapamycin on weight gain in female C57BL/6 mice on HFD started at the age of 7.5 months. By the age of 16 and 23 months, mice on HFD weighed significantly more (52 vs 33 g; p = 0.0001 and 70 vs 38 g; p < 0.0001, respectively) than mice on low fat diet (LFD). The i.p. administration of 1.5 mg/kg rapamycin, 3 times a week every other week, completely prevented weight gain, whereas administration of rapamycin by oral gavash did not. Rapamycin given in the drinking water slightly decreased weight gain by the age of 23 months. In addition, metabolic parameters were evaluated at the age of 16 and 23 months, 6 and 13 days after last rapamycin administration, respectively. Plasma leptin levels strongly correlated with body weight, (P < 0.0001, r=0.86), suggesting that the difference in weight was due to fat tissue mass. Levels of insulin, glucose, triglycerides and IGF1 were not statistically different in all groups, indicating that these courses of rapamycin treatment did not impair metabolic parameters at least after rapamycin discontinuation. Despite rapamycin discontinuation, cardiac levels of phospho-S6 and pAKT(S473) were low in the i.p.-treated group. This continuous effect of rapamycin can be explained by prevention of obesity in the i.p. group. We conclude that intermittent i.p. administration of rapamycin prevents weight gain without causing gross metabolic abnormalities. Intermittent gavash administration minimally affected weight gain. Potential clinical applications are discussed.

Keywords:

• aging
• mTO
• rapalogs
• rapamycin
• senescence

Introduction

In a wide range of doses and frequencies of administration, rapamycin extends life span in mice^1-26 As summarized by Anisimov et al, ²⁶ rapamycin has been administered via intraperetoneal (ip) and subcutaneous (sc) injections, by oral gavage, in the drinking water and food (chow). It is difficult to compare the efficiency of these schedules, given a variety of murine strains used, including inbred, heterogeneous, progeric and cancer-prone mice. The emerging pattern is that the life-extension by rapamycin is more prominent at higher doses and frequencies.^15,14,26 Yet in high-dose daily administrations, rapamycin will unlikely be eagerly used to suppress aging in humans due to concerns of side effects. Also, in theory, pulse-treatment (intermittent treatment) may improve tissue regeneration.²⁷ Therefore, comprehensive comparison of doses and frequencies is needed for the development of effective and simultaneously safe anti-aging strategies.

Obesity, the most common age-related condition, accelerates aging and age-related diseases.^28-32 It was shown that rapamycin decreases obesity in rodents and humans.^33-41 If used properly, this “immunosuppressant” can stimulate immunity.⁴²

In some studies, rapamycin caused insulin resistance. ^33,38,43 These metabolic changes are reversible.⁴⁴ Like so called “starvation-diabetes,” this condition may be benevolent and associated with longevity.⁴⁵ Also, “starvation-like” benevolent diabetes seems to be achieved only at high everyday doses and in some strains like C57BL/6NCr. In most other cases, rapamycin increases insulin sensitivity.^46-50

Here we sought to compare several routes of rapamycin administration using prevention of obesity on high fat diet (HFD) as a biomarker. In addition, we sought to investigate the effect of prolonged rapamycin treatment on glucose and lipid metabolism. To exclude direct effects of rapamycin, we measured these parameters after rapamycin discontinuation for 6-13 days (post-treatment levels). In addition, we compared rapamycin and metformin, which is known to improve metabolic parameters in humans.^32,51-55

Results

Rapamycin by i.p. injections prevented weight gain and obesity-associated leptin in female mice on HFD

Eighty 7.5 months old female mice of C57BL/6NCr strain were divided in 6 groups. One group received regular low fat (LF) chow (LF): 5% fat. Other 5 groups were fed high (60%) fat diet (HFD): 60% fat. These HFD groups included: untreated (control); R/ip group, which was treated with rapamycin via i.p. 3 times per week every other week; R/gavage group received rapamycin through gavage 3 times per week every other week (gav.); R/drinking group received rapamycin in drinking water (R/d.w) and Metformin (Met) group was given metformin in drinking water (Fig. 1). Treatment began at the same time as mice were put on high fat diet (HFD). Weight was measured weekly and metabolic parameters were evaluated at the age of 16 and 23 months, 6-13 days after rapamycin discontinuation. As shown in Figure 1A and B, at age of 16 and 23 months mice on HFD weighed significantly more than mice on regular diet (52 vs 33 g and 70 vs 38 g, respectively). Rapamycin administered via i.p. completely prevented weight gain on HFD (Fig. 1A and B). The body weight strongly correlated with blood levels of leptin, suggesting that the difference in weight was due to fat tissue mass (Fig. 1C and D). 15-month administration of rapamycin or metformin in drinking water resulted in slight but statistically significant decrease in weight gain by the age of 23 months in mice on high fat diet (Fig. 1B). Levels of insulin, glucose, triglycerides and IGF1 were not statistically different between all the groups, including treated and control mice on HFD and LFD (Figs. 2–4), indicating that prolonged administration of rapamycin did not have negative effects on metabolic parameters at the doses and schedules of treatment used. Notably, by the age 23 months (after 15-month treatment) mice in R/ip group showed statistically significant decrease in fasted insulin levels (p < 0.05), while glucose levels were similar to control mice, indicating their improved sensitivity to insulin (Fig. 3A).

Figure 1. Body weight and leptin levels of female mice treated with different schedules of rapamycin or metformin. Weight (grams) of 16-month old mice, 8 months after beginning of the treatments. Low (5%) fat (LF) diet (LF); all other groups received high (60%) fat diet (HFD): 60% fat. These HFD groups included: untreated (control); R/ip group, which was treated with i.p. rapamycin 3 times per week every other week; R/gavage group received rapamycin through gavage 3 times per week every other week (gav.); R/drinking group received rapamycin in drinking water (R/d.w) and Metformin (Met) group was given metformin in drinking water. (B) Weights of 23-month old survived mice, 15 months after the beginning of treatments. (C) Leptin (ng/ml) concentration was determined in fasted blood of 16-month old mice 8 months after beginning treatments. Fasted blood was collected in the morning after overnight fasting. (D) Correlation between leptin levels and weight in 16-month old mice (Data presented in A and C). Data presented as mean ± SE; p values are provided for observed differences in comparison with control (HFD) group; r – Pearson coefficient.

Figure 2. Metabolic profiles of 16 month-old mice on high fat diet 8 months from the beginning of the of treatment with different schedules of rapamycin or metformin. p values: the differences with control group (HFD). Glucose, insulin and triglycerides were determined in fasted serum. Data presented as mean ± SE. Fasted blood was collected in the morning after overnight fasting.

Figure 3. Blood glucose and insulin levels in 23 month-old mice on high fat diet 15 months from the beginning of the treatment with different schedules of rapamycin or metformin. Insulin (A and B) and glucose (C and D) concentrations were determined in fasting and non-fasting sera. Data presented as mean ± SE.

Prolonged treatment with rapamycin via i.p. resulted in sustained suppression of mTOR activity in murine hearts

At the age of 23 months (after 15-month treatment) mice were sacrificed 13 days after last rapamycin administration. As shown in Figure 5, levels of phospho-S6 and pAKT(S473) were low in the hearts of i.p. - treated mice compared with control (untreated mice on HFD) and gavage-treated group (Fig. 5).

Figure 4. Blood IGF1 and triglyceride levels in 23 month-old mice on high fat diet 15 months from the beginning of treatment with different schedules of rapamycin or metformin. IGF1 and triglyceride concentration. Data are mean ± SE.

Figure 5. Cardiac levels of p-S6 in mice on HFD. Immunoblot analysis of protein lysates from the heart of 23 month-old female mice on high fat diet: control – untreated; R/ ip – treated with rapamycin i.p 3 times a week every other week; R /gavage – treated with rapamycin through gavage 3 times a week every other week. Mice were sacrificed after overnight fasting, 13 days after last treatment.

Discussion

Here we demonstrated that i.p. injections of rapamycin prevented weight gain on high fat diet, whereas rapamycin by gavash (also given 3 times per 2 weeks) did not. Orally administrated (gavash) rapamycin has poor bioavailability. The i.p. route of its administration ensures high peak of systemic levels of rapamycin. We can conclude that acute high levels of rapamycin may be necessary to prevent obesity. Noteworthy the i.p. therapy was effective even given 3 times per week every other week.

The i.p. administration of 2 mg/kg rapamycin just once a week to HFD-fed C57BL/6 mice decreased weight gain and protected against insulin resistance.⁴¹

On the other hand, oral rapamycin in food daily did not prevent weight gain.⁴⁴

It may seem paradoxical that intermittent administration (by i.p.) is more effective than everyday administration (by oral gavash). One plausible explanation is that at high peak levels, rapamycin may affect cell types that are not sensitive to low concentrations of rapamycin. In fact, the effective concentrations of rapamycin vary broadly in cell lines in culture.

Although i.p. injections are not suitable for prolonged treatment in humans, we suggest that the development of highly bioavailable oral preparations of rapamycin are warranted for obesity treatment in humans. Alternatively, existing formulations of rapamycin could be used in high doses (as single dose) to reach desired peak concentrations in humans. Although high doses of rapamycin can cause side effects if used daily, intermittent administration (one a week or 3 times per 2 weeks, for example) may be sufficient for anti-obesity effects.

In our study, chronic administration of rapamycin in drinking water, which was previously shown to extend life-span in mice, slightly decreased weight gain after 15 months of treatment and this effect was comparable with the effect of metformin. Importantly, levels of glucose and insulin were the same in all groups in the post-treatment period. We measured these parameters 6-13 days after last rapamycin administration (post treatment) to avoid direct effect of rapamycin on metabolism. It was important to evaluate the internal metabolic health in post treatment period. We found no alterations by rapamycin treatment at all schedules. Levels of triglycerides and IGF1 were also unchanged. Unexpectedly, levels of p-S6 were decreased in the i.p. group even after rapamycin discontinuation. This seemingly puzzling result can be explained by the prevention of obesity in the i.p. group. It was shown that obesity^56-59 and aging^60,61 can be associated with increased fasting p-S6 as a marker of increased mTOR activity. This suggests that prevention of obesity by rapamycin may decrease basal level of mTOR for a prolong time.

Rapamycin (and other agents and conditions that inhibit mTOR) suppress gero-conversion from quiescence to senescence.^62-70 In the organism, suppression of geroconversion prevents age-related diseases in mammals^70-78 and pathological conditions in invertebrates, ^79,80 extending healthy lifespan.^81-83

Methods

Mice

Animal studies were conducted in accordance with the regulations of the Committee of Animal Care and Use at Roswell Park Cancer Institute. 80 female mice of C57BL/6NCr strain (7.5 month old) were divided into groups (10 mice per group and 2 groups: 20 mice per group): one group received standard laboratory chow (5% fat, Regular Diet – RD or low fat diet LFD). All other groups were fed on 60% fat chow (High fat diet - HFD) (Reaserch Diets, Inc., Cat # D12492 Rodent Diet 60% kCal fat; New Brunswick, NJ, USA). These groups on HFD were as follows: control – untreated, R/ ip group (N = 10 initially) was treated with rapamycin (LC Laboratories, Woburn, MA) via i.p. injections 1.5 mg/kg 3 times a week every other week; R/gavage group received 1.5 mg/kg rapamune (Sirolimus, rapamycin) (Cardinal Health, Syracuse, NY, USA) as gavage 3 times a week every other week; R/ drinking group received rapamycin in drinking water; Metformin group received metformin(Sigma-Aldrich, St. Louis, MO, USA) in drinking water.

Mice were sacrificed, if they lose weight more than 15% (according to the institution guidelines) or developed sickness such as dermatitis.

At 8 months from the beginning of treatment mice (reached 16 months of age) were fasted overnight and fasting blood was collected for biochemical analysis. After 15 months of treatment, mice (reached 23 months of age) were fasted overnight and sacrificed 13 days after the last treatment. Blood was collected at the end of the day before food was removed for overnight fasting. Next morning, fasted blood was collected. Non-fasted and fasted sera were prepared for biochemical analysis.

Glucose concentration in blood sera was measured using Accu-Chek Aviva strips (MaKesson, Atlanta, GA).

Insulin, IGF1, leptin and triglyceride concentration in sera were measured using Insulin (Mouse) Ultrasensitive ELISA kit (Alpco Diagnostics, Salem, NH), IGF1 (Mouse/Rat) ELISA kit (Alpco), Mouse leptin ELISA kit (Crystal Chem Inc., Downers Grove, IL) and Triglyceride Colorimetric Assay kit (Cayman Chemical Company, Ann Arbor, MI), respectively, as previously described ^23,35,65,84

Data were analyzed using range of standards and 4 parameter logistic fit or linear regression.

Statistical analysis

T test and correlation analyses (Pearson r coefficient and p value (2-tailed) were performed using GraphPad Prism version 6 for Windows, GraphPad Software (San Diego, CA; www.graphpad.com).

Immunoblot analysis

Tissues were homogenized and immunoblotting was performed as previously described.^23,35,65,84

Rabbit anti-phospho-S6 (Ser 240/244), anti-phospho-AKT(Ser 473) and anti-S6 were purchased from Cell Signaling Biotechnology (Danvers, MA).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.