Increasing Dietary Leucine Intake Reduces Diet-Induced Obesity and Improves Glucose and Cholesterol Metabolism in Mice via Multimechanisms

Male C57BL/6J mice were purchased from The Jackson Laboratories. Animal protocols were in compliance with the accepted standards of animal care and were approved by the Columbia University institutional animal care and use committee. Mice were maintained at 22°C on a 12:12 light-dark cycle (0700–1900 h) and had ad libitum access to a regular rodent chow (Purina Rodent Diet 20 5053) or, when indicated, a HFD containing 60% fat calories (D12492; Research Diets, New Brunswick, NJ). Both diets contained 20% protein calories. Leucine was supplemented via drinking water made of 1.5% (wt/vol) L-leucine (Sigma, St. Louis, MO).

Blood glucose levels were measured in tail-vein blood using a Glucometer Elite (Bayer, Elkhart, IN). For glucose tolerance tests, blood glucose was measured at 0, 30, 60, and 90 min after a bolus intraperitoneal glucose administration (1 mg/g body wt) to overnight-fasted mice. For insulin tolerance tests, regular human insulin was administered intraperitoneally (0.75 mU/g) to mice after 4-h food and leucine deprivation, and blood glucose was measured at 0, 30, 60, and 90 min after insulin injection. The homeostasis model assessment of insulin resistance (HOMA-IR) index was calculated based on the conventional formula: HOMA-IR = basal glucose (mmol/l) × basal insulin (mU/l)/22.5.

Five-month-old male C57BL/6J mice that had been on a HFD with or without leucine supplementation for 10 weeks were implanted with E-mitters (Mini-Mitter) for temperature measurements under anesthesia and were allowed a 4-day recovery before being housed in metabolic chambers. Oxygen consumption, locomotor activity, core temperature, and respiration exchange ratio were obtained continuously during 12:12 light-dark cycles using a comprehensive laboratory animal monitoring system (Columbus Instruments, Columbus, OH) open-circuit indirect calorimetry system. The same diet and leucine regimens for the mice were continued during indirect calorimetry. Data were collected over at least 5 days following 2–4 days of adaptation to the metabolic cages.

Quantitative real-time RT-PCR was used to determine mRNA expression as previously described (27,28). Briefly, total RNA was isolated using miniRNeasy columns (Qiagen, Valencia, CA) and reverse transcribed into single-stranded cDNA using random hexamers and M-MLV Reverse Transcriptase (Invitrogen, Carlsbad, CA). Quantitative amplification by PCR was carried out using Bio-Rad iQ SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA). Cyclophilin A was used as a control. Cycle numbers at a set fluorescence threshold within the exponential amplification range (C_t) were used to calculate expression levels of the genes of interest (relative to cyclophilin A). Changes in gene expression were expressed as fold increase or decrease over a specified reference sample. The following primers were used in this study: uncoupling protein (UCP)1, 5′-ggtgagttcga-caacttccg-3′ (forward) and 5′-ggtgatggtccctaa-gacacc-3′ (reverse); UCP2, 5′-gtggtcggagataccagagc-3′ (forward) and 5′-gaggttggctttcaggagag-3′ (reverse); UCP3, 5′-caagggagcggaccactcc-3′ (forward) and 5′-ctctctcctccagttcccatg-3′ (reverse); and cyclophilin A, 5′-atggcactggcggcaggtcc-3′ (forward) and 5′-ttgccattcctggacccaaa-3′ (reverse).

Gastrocnemius muscle (30 mg) or adipose tissues (100 mg) were powdered in liquid nitrogen and homogenized in a homogenization buffer (50 mmol/l Tris-HCl, pH 8.0, 1% NP-40, 1 mmol/l EDTA, 150 mmol/l NaCl₂, 50 mmol/l NaF, 0.5 mmol/l NaVO4, 5 mmol/l Na4P2O7, and 1 mmol/l DTT and protease inhibitors). Tissue homogenates were sonicated for 2 min in an ice-water bath, followed by centrifugation at 14,000 rpm for 15 min in an Eppendorf Microcentrifuge to remove insoluble cellular debris. Proteins (50–100 μg) were resolved by SDS-PAGE and then transferred to nitrocellulose membranes. Levels of specific proteins were determined by immunoblotting using enhanced chemiluminescence detection horseradish peroxidase reagents (Amersham Biosciences, Piscataway, NJ) and antibodies to mouse UCP1 and UCP3 (Sigma-Aldrich, St. Louis, MO) and β-actin (Santa Cruz Biotechnology, Santa Cruz, CA). Immunoblotting signals were quantified using the Quantity One program (Bio-Rad).

Body composition was determined using a Minispec TD-NMR Spectrometer (Bruker Optics). Unless otherwise indicated, plasma lipids, amino acids, and hormones, including insulin, leptin, glucagon, adiponectin, corticosterone, and catecholamines, were measured at the basal state (after a 5-h food and leucine deprivation) in HFD leucine and HFD control mice. Assays were done by the Mouse Phenotyping Center, Vanderbilt University. Plasma hormone concentrations are shown in Table 1.

The effects of leucine supplementation on weight, body composition, food intake, and blood glucose levels in HFD-fed and chow-fed mice were assessed separately using one-way ANOVA (Statistica version 6; StatSoft, Tulsa, OK); Newman-Keuls tests were used for post hoc comparisons. Multiple regression analysis was used to assess direct effects of leucine versus leucine-mediated reduction in adiposity on glucose and cholesterol metabolism. A two-tailed P < 0.05 was considered statistically significant. For indirect calorimetry measurements, repeated-measures ANOVA (Prism; GraphPad Software, San Diego, CA) were used. Post hoc analyses were conducted using Tukey post hoc or Bonferroni comparisons. Data are expressed as means with SEs.

Eight week-old male C57BL/6J mice on a regular rodent chow diet (chow) or on a HFD were supplemented with leucine via drinking water (chow-Leu and HFD-Leu, respectively). HFD and leucine supplementation were started simultaneously. Control mice on the respective diets were provided with regular tap water (chow-W or HFD-W). The amounts of regular or leucine-supplemented water consumed by the mice were the same. Leucine-supplemented mice ingested an additional ∼55 mg leucine daily via water consumption, which nearly doubled the total daily leucine intake from food (∼56 mg from chow; ∼63 mg from HFD) (Fig. 1A). Leucine supplementation increased plasma leucine concentrations by ∼30% during feeding but decreased overall plasma amino acid levels, particularly Asp, Glu, Gln, and Gly, in postabsorptive states and Asn, Pro, and Lys during feeding (Fig. 1 C and D), suggesting a decreased protein catabolism in leucine-supplemented mice.

Mice on the HFD gained significantly more weight, mainly as fat, than did those on the chow diet as was expected (Fig. 1,E and F). However, this diet-induced weight gain and expansion of fat mass were significantly attenuated in leucine-supplemented HFD-fed mice (Fig. 1,E and F). Eight weeks into the diet/supplement regimens, HFD-W mice weighed 41.4 ± 1.2 g, whereas HFD-Leu mice weighed 37.6 ± 1.0 g. This represents a 32% (P < 0.05) reduction of HFD-induced weight gain (12.1 g in HFD-W vs. 8.3 g in HFD-Leu, P < 0.05) (Fig. 1E) and a 25% decrease in adiposity (41.3% fat in HFD-W vs. 31.0% fat in HFD-Leu, P < 0.01). Differences in body weight, fat mass, and adiposity persisted throughout the 14-week study (Fig. 1,E and F). Leucine supplementation had no effect on body weight or composition in mice on the chow diet (Fig. 1,E and F). Total caloric intake was increased by 8.4% in chow-Leu mice, compared with chow-W mice, but was not significantly different between HFD-Leu and HFD-W mice (Fig. 1B).

Since weight loss without decreased food intake (the HFD group) and increased food intake without weight gain (the chow group) both suggested a leucine-mediated increase in energy expenditure, we directly measured energy metabolism in four pairs of HFD-W and HFD-Leu mice by indirect calorimetry. Total energy expenditure (24-h Vo₂) normalized to body weight was markedly increased (∼15%, P < 0.01) in HFD-Leu mice relative to HFD-W mice (Fig. 2A), despite a decrease in locomotor activity in HFD-Leu mice (Fig. 2B). Core body temperature was not significantly different between HFD-Leu and HFD-W mice (Fig. 2C). The respiratory exchange ratio (Vco₂/Vo₂) was slightly lower in HFD-Leu mice than in HFD-W mice (an average 0.808 vs. 0.814, respectively, P < 0.01) (Fig. 2D), suggesting increased fat oxidation in HFD-W mice.

To gain insight into molecular mechanisms for increased resting energy expenditure in HFD-Leu mice, we examined mRNA and/or protein levels of UCP1, -2, and -3, three major uncoupling proteins, in skeletal muscle and in brown adipose tissue (BAT) and white adipose tissue (WAT). Messenger RNA levels for UCP3, a major uncoupling protein in skeletal muscle and BAT, were increased by 1.51-fold (P < 0.05) in gastrocnemius muscle, 2.08-fold (P < 0.001) in perigonadal adipose tissue (WAT), and 1.67-fold (P < 0.05) in BAT of HFD-Leu mice, relative to the respective tissues of HFD-W mice (Fig. 3A). At the protein level, UCP3 was increased by 22.7% (P < 0.05), 28.9% (P < 0.01), and 20.1% (P < 0.05) in gastrocnemius muscle, perigonadal adipose tissue, and BAT, respectively, in HFD-Leu mice compared with control tissues (Fig. 3B). However, the levels of the ubiquitously expressed UCP2 and BAT UCP1 were not significantly different (Fig. 3 A and B). These results suggest that leucine may specifically regulate UCP3 gene expression in skeletal muscle, BAT, and WAT and that increased UCP3 expression may account, at least in part, for the increased resting energy expenditure in HFD-Leu mice.

Four weeks after HFD treatment, fasting blood glucose levels became elevated in HFD-W mice (112.5 ± 7.0 mg/dl) compared with chow-W mice (74.2 ± 3.6 mg/dl, P < 0.001) (Fig. 4A), but leucine supplementation largely prevented this HFD-induced hyperglycemia (89.2 ± 7.6 vs. 112.5 ± 7.0 mg/dl, P < 0.05). The glucose-lowering effect of leucine persisted throughout the study period, and HFD-Leu mice remained euglycemic at week 10 (Fig. 4A). Basal plasma insulin concentrations were significantly decreased in HFD-Leu mice compared with HFD-W mice (Table 1 and Fig. 4E). Glucose tolerance tests and insulin tolerance tests performed at weeks 10 and 11, respectively, confirmed that HFD-Leu mice were significantly more glucose tolerant and insulin sensitive than HFD-W mice (Fig. 4,B and C). The HOMA-IR index was >50% lower in HFD-Leu mice than in HFD-W mice (34.2 ± 7.6 vs. 67.7 ± 11.7, P < 0.05) (Fig. 4D). A forward stepwise regression analysis shows that leucine treatment (F = 5.6, P < 0.05) is a better predicting factor than adiposity (F = 3.5, P = 0.09) for HOMA-IR, suggesting that the glucose-lowering effect of leucine may not be entirely dependent on the decrease in adiposity.

Basal plasma glucagon levels were decreased by ∼50% (P < 0.05) in HFD-Leu mice (23.9 ± 1.3 pg/ml) compared with HFD-W mice (52.3 ± 9.9 pg/ml) (Fig. 4E and Table 1). Messenger RNA levels for glucose-6-phosphatase (G6P), a key enzyme in regulating hepatic glucose production, were also significantly lower in the livers of HFD-Leu mice, although mRNA levels for PEPCK and PGC-1α (peroxisome proliferator–activated receptor-γ coactivator), another two genes involved in hepatic gluconeogenesis (29,30), were not significantly different (Fig. 4F). G6P mRNA levels were significantly correlated with plasma glucagon levels (r = 0.85, P < 0.01) but not with insulin levels in HFD-fed mice (data not shown). Unlike HFD-fed mice, chow-fed mice remained euglycemic over the 14-week study period, and leucine supplementation had no additional effect on these values (Fig. 4A).

Leucine supplementation for 14 weeks decreased total cholesterol levels by 27% (P < 0.001) and LDL-cholesterol by 53% (P < 0.001) in HFD-fed mice (Fig. 5A). The cholesterol-reducing effect of leucine remained significant when adiposity was also included as an independent variable in a multiple regression analysis (Fig. 5B). This finding contrasts with the effect of leucine on plasma leptin levels. Leucine supplementation lowered plasma leptin levels (18.2 ± 5.4 ng/ml in HFD-Leu mice vs. 36.3 ± 3.3 ng/ml in HFD-W mice, P < 0.01) (Table 1), but this effect was essentially lost when the decrease of adiposity was taken into account (Fig. 5B). No significant differences were detected in HDL cholesterol, total triglyceride, or free fatty acid levels (Fig. 5A). Taken together, these data suggest that leucine has a selective effect on LDL cholesterol reduction, independent of its effect on adiposity.

We have shown that doubling dietary leucine intake substantially reduced diet-induced weight gain and improved glucose and cholesterol metabolism in mice. The attenuation of HFD-induced weight gain by leucine supplementation resulted from increased resting energy expenditure associated with increased UCP3 protein expression in skeletal muscle, BAT, and WAT. Food intake was not reduced by the increased leucine intake. Both we and others have observed that mice on chronic HFD become highly fuel efficient; they gain more weight per unit of calorie consumed compared with chow-fed mice (31–34). In HFD-Leu mice, increased UCP3-mediated mitochondrial uncoupling of oxidative phosphorylation in thermogenic tissues likely accounts for the increased resting energy expenditure with reduced fuel efficiency. Our result is consistent with the reports that increased UCP3 expression reduces HFD-induced weight gain in transgenic mice (35) and is associated with the obesity-resistant phenotype in rats (34). Additionally, increases in muscle UCP3 expression level are generally seen in obese rats and humans (36,37), and weight loss results in decreases in muscle UCP3 expression in humans (37,38). However, HFD-Leu mice, while having lower adiposity relative to HFD-W mice, express higher levels of UCP3 protein in both skeletal muscle and adipose tissues. Thus, the effect of increased leucine intake on tissue UCP3 expression appears to be independent of the attenuated adiposity in HFD-Leu mice. How leucine increases UCP3 expression is unclear, it may depend on leucine-mediated activation of the mTOR-S6K pathway, since insulin also both activates mTOR-S6K signaling pathway and increases UCP3 expression in skeletal muscle (39).

It is of interest to note that while increased UCP3 expression in skeletal muscle and in adipose tissues is consistent with the observed increased in resting energy expenditure, core body temperatures were not significantly different between HFD-Leu and HFD-W mice. Normal core body temperatures were associated with normal diurnal fluctuations in both HFD-Leu and HFD-W mice, suggesting that the thermoregulation mechanism was intact and that excess heat generated by increased energy expenditure was effectively dissipated in HFD-Leu mice. A similar result has also been observed in mice with genetic ablation of MCH (melanin-concentrating hormone), a hypothalamic neuropeptide involved in regulating food intake and energy expenditure; a 20% increase in the rate of oxygen consumption in MCH^−/− mice has no effect on core body temperature (40).

Cota et al. (10) showed that centrally administered leucine activates the mTOR pathway in the hypothalamic arcuate nucleus and suppresses food intake in rats. Orally administered leucine in our study, however, had no significant suppressive effect on food intake. Possible explanations for the apparent discrepancy include the doses and routes by which leucine was introduced into animals (intracerebroventricular infusion versus oral administration). Orally administered leucine may be substantially metabolized in the liver before it reaches other tissue targets. Plasma leucine concentrations were increased by 30% during feeding in HFD-Leu mice and were not significantly different in the postabsorptive state. Thus, it is possible that leucine concentrations in the hypothalamus of HFD-Leu mice did not reach a level comparable with that achieved by intracerebroventricular leucine infusion reported by Cota et al. However, this does not preclude the possible involvement of the central nervous system in mediating leucine's effects on energy expenditure, whether in animals receiving centrally administered leucine or in animals with increased oral intake of leucine, such as HFD-Leu mice. Cota et al. did not report the status of energy expenditure, but it is possible that changes in centrally mediated energy expenditure may require a lower central nervous system level of leucine than what is necessary for suppressing food intake. Another potential explanation for the apparent discrepancy between our study and that of Cota et al. is the species difference. Mice (ours) and rats (Cota et al.) may have species-specific difference in ways in which they respond to leucine.

Locomotor activity was decreased in HFD-Leu mice compared with HFD-W mice. The cause for this change is unclear. However, this change was not due to toxicity from leucine supplementation. Leucine is among the most tolerated amino acids as no adverse effects of increased leucine intake (usually three times the daily requirement) were reported in various animal and human studies (16,41). In our study, leucine-supplemented mice consumed normal amounts of food and water and had a normal growth curve compared with controls. In fact, on the rodent chow diet group, leucine-supplemented mice consumed more food than control mice. Thus, the decreased locomotor activity does not appear to be related to food-seeking activity either.

Increasing dietary leucine intake significantly improves glucose metabolism in HFD-fed mice. It appears that better glycemic control is primarily due to the improved insulin sensitivity that is associated with decreased adiposity in HFD-Leu mice. However, the involvement of other mechanisms cannot be ruled out. In fact, multiple regression analyses suggest that leucine's effect on the HOMA-IR index was only partially accounted for by the reduced adiposity. Two important players in the regulation of hepatic glucose production, plasma glucagon concentrations and hepatic G6P expression levels, were both lower in HFD-Leu mice. Additionally, plasma concentrations of glycogenic amino acids, which provide substrates for hepatic gluconeogenesis in the postabsorptive state (42), were also lower in HFD-Leu mice. Taken together, these results suggest that hepatic glucose production may be decreased in HFD-Leu mice, which could contribute to the improved glycemic control.

Leucine has a biphasic effect on glucagon secretion in the perfused rat pancreas: stimulatory in low glucose states (3.3 mmol/l) but inhibitory in high glucose states (8.3 mmol/l) (43) and may suppress glucagon secretion indirectly by lowering concentrations of glucagon secretagogues such as glutamine, glycine, and alanine in HFD-Leu mice (44,45). Leucine may also affect glucose metabolism by stimulating insulin secretion (43–45). Although fasting insulin levels were lower in HFD-Leu mice, most likely secondary to the improved insulin sensitivity, it cannot be ruled out that leucine may transiently stimulate insulin secretion in HFD-Leu mice. A direct action of leucine in skeletal muscle and liver may involve mTOR-mediated feedback inhibition of insulin signaling to affect glucose disposal and glucose production, respectively (17–20). However, our data show that increasing dietary leucine intake resulted in a net improvement of whole-body insulin sensitivity and glycemic control, along with reduced adiposity in mice on a chronic HFD.

Leucine markedly reduced plasma total and LDL cholesterol levels, independent of weight reduction. Cholesterol-reducing effects of L-arginine and keto and amino acid supplementation in diabetic rats and in patients on low protein diets have also been reported by others (46,47). Further elucidation of leucine signaling, via mTOR and/or other potential mediators, will be required to uncover the relevant molecular mechanisms.

Finally, despite the remarkable effects of leucine observed in HFD-fed mice, increased leucine intake had no notable effects on chow-fed mice, except that leucine-supplemented mice had slightly increased food intake. Since the increased food intake in these mice was not associated with increases in body weight, adiposity, or blood glucose levels, we speculate that this behavioral change was only secondary to a leucine-induced increase in basal metabolic rates, as a homeostatic mechanism to prevent downward excursion from the “normal” body weight.

In summary, we have shown that doubling dietary leucine intake on a chronic basis produces a net health benefit that includes reduction of diet-induced weight gain, hyperglycemia, and hypercholesterolemia in mice on a HFD. The reduction in adiposity results from a leucine-induced increase in resting energy expenditure and is associated with increased UCP3 protein expression in skeletal muscle and adipose tissues. The overall beneficial effects on lipid and glucose metabolism produced by leucine supplementation appear to involve multiple mechanisms that need to be further defined.

This study was supported by grants from National Institutes of Health, DK063034 (to Y.Z.), DK071866 (to Y.Y.), and DK066618 (to G.J.S.) and grants P30DK026687 and P30DK063068.

We thank Dr. Rudy Leibel for many discussions during the study and Drs. Rudy Leibel, Henry Ginsberg, and Ira Goldberg for reviewing the manuscript.