Reduced Liver Mitochondrial Energy Metabolism Impairs Food Intake Regulation Following Gastric Preloads and Fasting

Abstract

Objective: The capacity of the liver to serve as a peripheral sensor in the regulation of food intake has been debated for over half a century. The anatomical position and physiological roles of the liver suggest it is a prime candidate to serve as an interoceptive sensor of peripheral tissue and systemic energy state. Importantly, maintenance of liver ATP levels and within-meal food intake inhibition is impaired in human subjects with obesity and obese pre-clinical models. Previously, we have shown decreased hepatic mitochondrial energy metabolism (i.e., oxidative metabolism & ADP-dependent respiration) in male liver-specific, heterozygous PGC1a mice results in increased short-term diet-induced weight gain with increased within meal food intake. Herein, we tested the hypothesis that decreased liver mitochondrial energy metabolism impairs meal termination following nutrient oral pre-loads.

Methods: Liver mitochondrial respiratory response to changes in ΔG_ATP and adenine nucleotide concentration following fasting were examined in male liver-specific, heterozygous PGC1a mice. Further, food intake and feeding behavior during basal conditions, following nutrient oral pre-loads, and following fasting were investigated.

Results: We observed male liver-specific, heterozygous PGC1a mice have reduced mitochondrial response to changes in ΔG_ATP and tissue ATP following fasting. These impairments in liver energy state are associated with larger and longer meals during chow feeding, impaired dose-dependent food intake inhibition in response to mixed and individual nutrient oral pre-loads, and greater acute fasting-induced food intake.

Conclusion: These data support previous work proposing liver-mediated food intake regulation through modulation of peripheral satiation signals.

Keywords: ATP; Liver; fasting; food intake; mitochondria; satiation.

Figure 1.. LPGC1a+/− LPGC1a+/− Mice Have Reduced ATP Homeostasis During Fasting.

A) Respiration of isolated liver mitochondria during changes in ATP free energy (ΔG ATP) via creatine kinase clamp. B) Liver ATP, ADP, & AMP concentration (nmol/g) in mice after 2hr food withdrawal, 18 hr fast, or 4 hr refeeding. Impact of fasting/re-feeding on the liver was assessed as C) liver energy charge (ATP + (0.5 x ADP))/(ATP+ADP+AMP), D) NAD+/NADH, and mRNA expression of E) PGC1a, F) PGC1b, G) PPRC1, H) NT-PGC1a, I) Cox4i2, J) PPAR alpha, K) HADHA, L) HMGCS2, M) PPAR delta, and N) PPAR gamma in 2hr food withdrawn, 18hr fasted, or 4 hr refed mice. Data are represented as mean ± SEM. n=3-8 biological replicates for each genotype for adenine nucleotide concentration. n=8-10 biological replicates for each genotype for all other data. # main effect fasting or refeeding compared to 2hr, & main effect of genotype, + main effect of refed vs fasted by two-way ANOVA. *p<0.05 between genotypes by Student’s t-test (A). # main effect vs 2hr, + main effect refed vs fasted, & main effect LPGC1a+/− vs WT. ## fasting versus 2 hr within genotype, && LPGC1a+/− versus wildtype within group, and ++ refed versus fasting within genotype pairwise comparisons were performed using Fishers LSD. See also Figure S2 for liver mRNA expression of mitochondrial in 2hr food withdrawn, fasted, and 4 hr refed mice.

Figure 2.. Reduced Liver-Specific PGC1a Results in Underlying Differences in Feeding Behavior.

A) Acute basal food intake at the beginning of the dark cycle following 2hr food withdrawal. B) Average number of meals, C) grams per meal, and D) length of each meal during each 1hr period. Data are represented as mean ± SEM. n = 13 technical replicates and n= 7 – 9 biological replicates for each genotype. *p < 0.05 between genotypes within time by Student’s t test.

Figure 3.. Dose Dependent Impairment in Satiation Following Mixed Macronutrient Oral-Preloads in LPGC1a+/− Mice.

Percent inhibition of food intake following oral pre-load of Ensure^® at A) 5, B) 10, & C) 15 mL/kg and D) 10 mL/kg 1% methylcellulose. Average duration of meal per 1 hr period following mixed nutrient E) 5, F) 10, & G) 15 mL/kg, and H)10 m/kg methylcellulose oral pre-load. Average meal size period 1 hr period following I) 5, J) 10, & K) 15 mL/kg nutrient, or L) 10 mL/kg non-nutritive oral pre-load. Ensure^® oral pre-load food intake inhibition dose response at M) 1 hr, N) 2 hr, & O) 3 hr. Data are represented as mean ± SEM. n= 7 – 9 biological replicates for each genotype. *p < 0.05 between genotypes within time by Student’s t test. See also Figure S4 for meals per period data across the 3 oral pre-load doses.

Figure 4.. Decreased Satiation Response in LPGC1a+/− Mice Following Oral Delivery of Individual Macronutrients.

Percent inhibition of food intake following oral pre-load of individual macronutrients at 10 mL/kg: A) 40% Glucose, B) Intralipid^®, & C) Proteinex^®. Average duration of meal per 1 hr period following D) 40% Glucose, E) Intralipid^®, & F) Proteinex^® individual macronutrient oral pre-load. Average meal size per 1 hr period following G) 40% Glucose, H) Intralipid^®, & I) Proteinex^® nutrient oral pre-load. Data are represented as mean ± SEM. n= 7 – 9 biological replicates for each genotype. *p < 0.05 between genotypes within time by Student’s t test. See also Figure S5 for meals per period data across the 3 individual macronutrient oral pre-loads.

Figure 5.. Peripheral Satiation Signal Receptor Inhibition Does Not Impact Food Intake Following Oral Pre-Loads in LPGC1a+/− Mice.

Percent increase in chow intake during acute mixed nutrient oral pre-load (10 mL/kg) following intraperitoneal delivery of satiation signal receptor inhibitors A) Ondanestron (5HTR3 antagonist, 5 mg/kg), B) Lorglumide (CCK1R antagonist, 300 μg/kg), and C) Exendin-9 (GLP1R antagonist, 100 μg/kg). Data are represented as mean ± SEM. n= 7 – 9 biological replicates for each genotype. *p<0.05 between genotypes within time by Student’s t test.

Figure 6.. Fasting Produces Greater Increases in Acute Food Intake in LPGC1a+/− Mice.

A) Cumulative food intake, B) grams/meal per 0.5 hr period, C) meals per period, and D) seconds per meal in WT and LPGC1a+/− mice fasted for 18 hr compared to 2hr food withdrawal. E) Cumulative food intake and F) percent inhibition of food intake in 18hr fasted mice receiving 10 μg/kg CCK-8S prior to access to food. Data are represented as mean ± SEM. n= 7 – 9 biological replicates for each genotype. *p<0.05 between genotypes within time and # p<0.05 between treatment within genotype by Student’s t test.

Figure 7.. Maladaptive Response of Hypothalamic POMC and AgRP Expression to Fasting in LPGC1a+/− Mice.

Gene expression of A) POMC, B) AgRP, and C) NPY in the hypothalamus of 2hr food withdrawn, fasted, and 4 hr refed mice. In situ hybridization of D) POMC and E) AgRP in the arcuate nucleus of the hypothalamus of 2hr food withdrawn wildtype and LPGC1a+/− mice. Data are represented as mean ± SEM. n= 8 – 10 biological replicates for each genotype. # main effect fasting compared to 2hr by two-way ANOVA (B). ## fasting versus 2 hr within genotype, && LPGC1a+/− versus wildtype within group, and ++ refed versus fasting within genotype pairwise comparisons were performed using Fishers LSD (A, B, & C). D) and E) genotyped compared by Student’s t test. Representative fluorescent in situ hybridization images from analysis of the total area of POMC and AgRP expression are presented in Figure S8.