Low protein-induced increases in FGF21 drive UCP1-dependent metabolic but not thermoregulatory endpoints

Abstract

Dietary protein restriction increases adipose tissue uncoupling protein 1 (UCP1), energy expenditure and food intake, and these effects require the metabolic hormone fibroblast growth factor 21 (FGF21). Here we test whether the induction of energy expenditure during protein restriction requires UCP1, promotes a resistance to cold stress, and is dependent on the concomitant hyperphagia. Wildtype, Ucp1-KO and Fgf21-KO mice were placed on control and low protein (LP) diets to assess changes in energy expenditure, food intake and other metabolic endpoints. Deletion of Ucp1 blocked LP-induced increases in energy expenditure and food intake, and exacerbated LP-induced weight loss. While LP diet increased energy expenditure and Ucp1 expression in an FGF21-dependent manner, neither LP diet nor the deletion of Fgf21 influenced sensitivity to acute cold stress. Finally, LP-induced energy expenditure occurred even in the absence of hyperphagia. Increased energy expenditure is a primary metabolic effect of dietary protein restriction, and requires both UCP1 and FGF21 but is independent of changes in food intake. However, the FGF21-dependent increase in UCP1 and energy expenditure by LP has no effect on the ability to acutely respond to cold stress, suggesting that LP-induced increases in FGF21 impact metabolic but not thermogenic endpoints.

Figure 1

Low protein-induced increases in energy expenditure require UCP1. Wildtype and Ucp1-KO mice were placed on control or LP diet (8/group) at 28 °C, with energy expenditure being recorded for the first 6 days on diet. Raw oxygen consumption (VO2) following transition to control or LP diet in wildtype (A) and Ucp1-KO (B) mice. Average EE (C), EE normalized to body weight (D), RER (E), and activity (F) over days 5 and 6 of the experiment. *P < 0.05 LP vs. respective control.

Figure 2

UCP1 deletion enhanced low protein-induced weight loss in part by reversing the effect of low protein on food intake. Body weight change (A), average daily food intake (B), fat (C) and lean mass (D) gain in wildtype and Ucp1-KO mice consuming control or LP diet (8/group) for 6 weeks at 28 °C. *P < 0.05 LP vs. respective control. ^#P < 0.05 WT-LP vs. Ucp1-LP

Figure 3

Effect of low protein on FGF21 and liver and BAT metabolic gene expression in Ucp1-KO mice. Tissues were collected from wildtype and Ucp1-KO mice after consuming control or LP diet (8/group) for 6 weeks. Serum FGF21 was measured via ELISA (A), while liver Fgf21 (B,C) and BAT (D) gene expression was measured via real-time PCR. *P < 0.05 LP vs respective control; ⁺P < 0.05 WT-CON vs. Ucp1-CON; ^#P < 0.05 WT-LP vs. Ucp1-LP. Probability values for the two-way ANOVA main effects of diet and genotype, and their interaction (G*D) are provided below the respective mRNA in C and D.

Figure 4

Low protein-induced FGF21 increases EE but does not alter the response to acute cold stress. WT and Fgf21-KO mice were placed on LP diet for 10 days (10/group) at room temperature, with body weight change (A) and rectal temperature (B) measured every two days. On day 10 food was removed and temperature was reduced to 4 °C for 6hrs, with rectal temperature (C) and EE (D, E, F) measured. Raw EE (E) and EE normalized to body weight (F) for the 24 hr prior to cold exposure (23 °C) and the final 3 hr of cold exposure (4 °C). Gray shading in Panel D represents lights off, and the blue shading represents the period of cold exposure. *P < 0.05 LP vs. respective control.

Figure 5

Effect of low protein diet and FGF21 deletion on metabolic endpoints following acute cold stress. WT and Fgf21-KO mice were placed on LP for 10 days, and on day 10 food was removed and temperature was reduced to 4 °C. Mice were sacrificed after 6hrs at 4 °C, and serum FGF21 (A), liver Fgf21 mRNA (B), liver (C) and BAT (D) mRNA expression was measured. 10/group. *P < 0.05 LP vs. respective control. Probability values for the two-way ANOVA main effects of diet and genotype, and their interaction (G*D) are provided below the respective mRNA in C and D.

Figure 6

Hyperphagia is not necessary for low protein-induced increases in energy expenditure. Wildtype mice were placed on control or LP diet ad libitum for 7 days, or were placed on LP but pair-fed to the daily intake of the control group (LPPF) to prevent hyperphagia. Daily food intake (A,B), body weight on day 0 (C) and day 7 (D), and BW change for the 7 days (E). Unadjusted average daily energy expenditure (F) and averaged across days 5–7 (G) and EE adjusted for body weight on day 7 (H and I). 8 mice/group. *P < 0.05 vs. control. ^#P < 0.05 vs. LP.