The short-chain fatty acid acetate reduces appetite via a central homeostatic mechanism

Abstract

Increased intake of dietary carbohydrate that is fermented in the colon by the microbiota has been reported to decrease body weight, although the mechanism remains unclear. Here we use in vivo(11)C-acetate and PET-CT scanning to show that colonic acetate crosses the blood-brain barrier and is taken up by the brain. Intraperitoneal acetate results in appetite suppression and hypothalamic neuronal activation patterning. We also show that acetate administration is associated with activation of acetyl-CoA carboxylase and changes in the expression profiles of regulatory neuropeptides that favour appetite suppression. Furthermore, we demonstrate through (13)C high-resolution magic-angle-spinning that (13)C acetate from fermentation of (13)C-labelled carbohydrate in the colon increases hypothalamic (13)C acetate above baseline levels. Hypothalamic (13)C acetate regionally increases the (13)C labelling of the glutamate-glutamine and GABA neuroglial cycles, with hypothalamic (13)C lactate reaching higher levels than the 'remaining brain'. These observations suggest that acetate has a direct role in central appetite regulation.

Figure 1. FCs reduce food intake and increase faecal SCFA concentrations.

(a) Body weight gain and (b) average weekly food intake of mice fed with either a HFD supplemented with the highly fermentable fibre inulin (HF-I) or a relatively non-fermentable fibre cellulose (HF-C). Body weight gain and average food intake significantly reduced in the HF-I group, **P<0.01, **P<0.05 based on two-sided, unpaired Student’s t-test(n=12 per group). (c) Total and acetate-only faecal SCFA concentrations obtained from HF-C and HF-I fed mice **P<0.01 based on two-sided, unpaired Student’s t-test (n=6 per group randomly selected from the n=24 cohort). (d) The biodistribution of carbon-11 (¹¹C) i.v. acetate as determined using PET scanning. Image depicts a fasted mouse following i.v. infusion. Image shows uptake in the brain, liver and heart. (e–g) Brain, liver and heart uptake of i.v. and colon infused ¹¹C-acetate as expressed as a percentage of the initial dose administered. No significant differences, when compared by GEE and Mann–Whitney U-test, were observed between i.v. administrations in the fed or fasted state, but there is a slower increase when the ¹¹C-acetate was colonically infused (P<0.001; n=3–4 per group).

Figure 2. Acetate administration reduces food intake.

(a) Serum acetate concentrations of mice administered with 500 mg kg⁻¹ i.p. acetate or saline (n=12). (b) Acute food intake data showing food intake in mice following the acute i.p. administration of acetate (500 mg kg⁻¹) or saline. Food Intake significantly reduced at 0–1 and 0–2 h; ***P<0.001, **P<0.01 (n=21–22 per group). (c) The effect of 2.5 μmol of sodium acetate administered into the third ventricle of cannulated rats on food intake compared with a sodium chloride control injection. Cross-over data (n=15) compared using two-sided, paired Student’s t-test (P=0.05, P=0.08). (d) Behavioural response observed in mice administered either i.p. acetate, saline or lithium chloride (positive control). There was no significant effect on behaviour between saline or acetate as compared using Kruskal–Wallis non-parametric ANOVA (n=8). (e) Change in baseline blood glucose concentrations (fold change) following 500 mg kg⁻¹ sodium acetate or saline injection in ad libitum fed mice (treatment effect P=0.6 as determined by two-way ANOVA; n=9–10 per group). (f) The effect of i.p. liposome-encapsulated acetate on acute food intake. No significant difference throughout based on two-sided, unpaired Student’s t-test (n=7). (g,h) Effects of i.p. acetate on plasma concentrations of appetite-regulating gut peptides PYY and GLP-1. Measurements made 60 min post injection. There was no significant effect on either peptide as determined by two-sided, unpaired, Student’s t-test.

Figure 3. Acetate effects in the hypothalamus.

(a) Regions of interest (ROIs) used in manganese-enhanced MRI (MEMRI). MR image showing the ROI locations in the hypothalamus from which signal intensity (SI) measurements were determined. ARC, arcuate nucleus; VMH, ventromedial hypothalamus; ArP, area postrema; NTS, nucleus of solitary tract. White bar represents 1 mm. (b–e) Hypothalamic neuronal activation in the ARC, VMH, the PVN and the NTS of mice following i.p. administration of acetate or a saline control as determined by MEMRI. Arrow signifies start of i.v. MnCl₂infusion. SI is significantly increased in the ARC of acetate treated mice compared with saline-injected controls based on generalized estimated equations (GEE) and Mann–Whitney U-test; **P<0.01 (n=4–5 per group). (f) Effect of acetate on hypothalamic expression of POMC, NPY and AgRP as determined by hypothalamic qPCR. Data expressed as fold change in expression compared with saline-injected controls at 30 and 60 min post administration one-way ANOVA with post hoc Dunnett’s correction; **P<0.01, ***P<0.001 (n=5 per group). (g,h) pACC and pAMPK hypothalami content expressed in relation to β-actin control, based on two-sided, unpaired Student’s t-test; *P<0.05 (n=5). (i) Immunoblots of hypothalamic pAMPK and pACC levels in mice 30 min after the i.p. injection of either saline or acetate (full blot with annotation available in Supplementary Fig. 1). (j) Proposed mechanism of acetate induced inhibition of the feeding impulse. Relatively increased hypothalamic TCA cycle activity increases ATP production, decreases AMP levels and AMPK inhibition of acetyl-CoA carboxylase (ACC), increases malonyl-CoA levels and stimulates POMC mRNA expression, and inhibits appetite.

Figure 4. Effects of peripheral (2-¹³C) acetate or intragastric (U-¹³C) inulin administrations on hypothalamic and cerebral metabolism.

(a) Representative ¹³C (125.03 MHz) HR-MAS spectra (4 °C, 5 kHz) of the hypothalamus from a mouse fasted overnight, 15 min after i.p. [2-¹³C] acetate administration (500 mg kg⁻¹). Inset: Representative ¹³C HR-MAS spectra (28–38 p.p.m.) from the hypothalamus of an overnight-fasted mouse, 180 min after [U-¹³C] inulin administration (100 mg) by gavage. (b) Increases in ¹³C incorporation into the acetate C2, GABA C2, Glu C4, Gln C4 and Lac C3 carbons (mean+s.d.) in the hypothalamus and remaining brain biopsies, following 0, 15, 30 min i.p. [2-¹³C] acetate (n=18) or 180 min intragastric [U-¹³C] inulin administrations *P<0.05, **P<0.01, ***P<0.001 (n=4). (c) Summary of the effects of [2-¹³C] acetate or [U-¹³C] inulin administrations in hypothalamic metabolism. Extracellular [2-¹³C] or ¹³C₂ acetate derived from plasma or cerebrospinal fluid are transported primarily to the astrocytes, where they are metabolized oxidatively in the TCA cycle, labelling astrocytic glutamate C4 and glutamine C4 that exchange with the corresponding neuronal pools through the glutamate–glutamine cycle, labelling GABA only in gabaergic neurons. (2-¹³C) or ¹³C₂ acetate are also oxidatively recycled in the neuronal cycle, originating the Lac C3 resonance. The glutamate–glutamine and GABA cycles support glutamatergic and gabaergic neurotransmissions, the two fundamental synaptic events triggering increased Ca²⁺ uptake and competitive Mn²⁺ uptake (MEMRI) that determine the appetite impulse.

Figure 5. Relative changes in total metabolite content in the hypothalamus and remaining brain biopsies after (2-¹³C) acetate or (U-¹³C) inulin administrations.

(a) ¹HR-MAS spectrum from a representative biopsy from the hypothalamus (black) and superimposed LCModel fitting (red). (b–f) Relative ¹H HR-MAS ratios Ac/Ino, GABA/Ino, Glu/Ino, Gln/Ino and Lac/Ino at increasing times after (2-¹³C) acetate or (U-¹³C) inulin administrations. Lac, lactate; Ac, acetate;, Glu, glutamate; Gln, glutamine; GABA, γ-amino butyric acid; Cr+PCr, creatine+phosphocreatine; Ino, myo-inositol; n.r, not resolved; CRLB, Cranmer–Rao Lower Bound <20%. Bar graphs represent the mean and standard deviation. Statistical significance was evaluated using two-tailed unpaired Student t-test. “Remaining brain” accounts for cerebral extra-hypothalamic tissue. *P<0.05, **P<0.01.