PI3K signaling in the ventromedial hypothalamic nucleus is required for normal energy homeostasis

Abstract

Phosphatidyl inositol 3-kinase (PI3K) signaling in the hypothalamus has been implicated in the regulation of energy homeostasis, but the critical brain sites where this intracellular signal integrates various metabolic cues to regulate food intake and energy expenditure are unknown. Here, we show that mice with reduced PI3K activity in the ventromedial hypothalamic nucleus (VMH) are more sensitive to high-fat diet-induced obesity due to reduced energy expenditure. In addition, inhibition of PI3K in the VMH impaired the ability to alter energy expenditure in response to acute high-fat diet feeding and food deprivation. Furthermore, the acute anorexigenic effects induced by exogenous leptin were blunted in the mutant mice. Collectively, our results indicate that PI3K activity in VMH neurons plays a physiologically relevant role in the regulation of energy expenditure.

Fig. 1

Deletion of p110α in SF1 neurons increases sensitivity to diet-induced obesity. (a) Weekly body weight was measured in group housed male mice weaned on regular chow (n=12/genotype). (b) Body composition was measured in 15-week old male mice fed with regular chow (n=12/genotype). (c) Weekly body weight was measured in group housed male mice weaned on HFD (n=16 or 23/genotype). (d) Body composition was measured in 18-week old male mice fed with HFD (n=10/genotype). (e) Serum leptin levels were measured in 7-month old male mice at both fed and fasted conditions (n=6/genotype). Data are presented as mean ± SEM, and * P<0.05 and **P<0.01 between p110α^lox/lox/SF1-Cre mice and p110α^lox/lox mice.

Fig. 2

Deletion of p110α in SF1 neurons reduces energy expenditure. (a–b) Daily food intake was measured in 7-week old male mice with comparable body weight fed with regular chow (a) or with HFD (b) (n=11–15/genotype). (c–g) Seven-week old chow-fed male mice (n=11 or 16/genotype) were fed with HFD for 2 weeks and matched for body weight (p110α^lox/lox: 23.8±0.7 g vs p110α^lox/lox/SF1-Cre: 24.9±0.7 g, P=0.31), followed by metabolic analyses using the TSE metabolic chambers. Data are presented as mean ± SEM, and * P<0.05 and **P<0.01 between p110α^lox/lox/SF1-Cre mice and p110α^lox/lox mice.

Fig. 3

Deletion of p110α in SF1 neurons disrupts the thermogenic regulation in response to HFD feeding and fasting. (a–d) Six-month old chow-fed male mice (n=6/genotype) were matched for body weight (p110α^lox/lox: 37.7±1.4 g vs p110α^lox/lox/SF1-Cre: 37.4±1.0 g, P=0.90) and adapted to the TSE metabolic chambers. The mice were provided with HFD at 17:00 (2 hr prior to dark cycle), and metabolic parameters were monitored from 24 hr before the HFD feeding till 24 hr afterwards using the TSE metabolic chambers. Upper panels: temporal levels of O₂ consumption (a), CO₂ production (b) and heat production (c). The arrow indicates the beginning of HFD feeding. Lower panels: changes in O₂ consumption (a), CO₂ production (b) and heat production (c) between the 12-hr dark cycle before HFD feeding and the 12-hr dark cycle afterwards in p110α^lox/lox/SF1-Cre and p110α^lox/lox mice. (d) HFD intake during the 12-hr dark cycle in the TSE chambers. (e–f) The mice were maintained on HFD for 6 weeks and weekly body weight gain (e) and energy intake (f) were recorded. (g–h) Five-month old chow-fed male mice (n=6/genotype) were matched for body weight (p110α^lox/lox: 31.5±0.6 g vs p110α^lox/lox/SF1-Cre: 32.4±1.1 g, P=0.51) and fasted for 24 hr. The body weight loss was measured (g) and heat production was recorded using the TSE chambers (h). (i) Fourteen-week old chow-fed mice were fed with HFD for 3 weeks (mean body weight: p110α^lox/lox: 33.1±3.4 g vs p110α^lox/lox/SF1-Cre: 38.8±2.6 g, P=0.26), and BAT were collected after euthanasia. Messenger RNA levels of indicated BAT genes were quantified with real-time PCR (n=6 or 7/genotype). Data are presented as mean ± SEM, and * P<0.05 and **P<0.01 between p110α^lox/lox/SF1-Cre mice and p110α^lox/lox mice.

Fig. 4

Deletion of p110α in SF1 neurons blunts the anorexigenic effects of central leptin. Five-month old chow-fed male mice (n=4 or 6/genotype) were matched for body weight, and received saline (1 μl, i.c.v.) at 16:00 followed by leptin (6 μg in 1 μl saline, i.c.v.) 24 hr later. Leptin-induced reductions in food intake (a), meal size (b), meal frequency (c) and RER (d) were monitored using the TSE metabolic chambers. (e) Leptin-induced weight loss was measured. Data are presented as mean ± SEM, and * P<0.05 and **P<0.01 between p110α^lox/lox/SF1-Cre mice and p110α^lox/lox mice.