Maternal low-calorie sweetener consumption rewires hypothalamic melanocortin circuits via a gut microbial co-metabolite pathway

Abstract

The prevalence of obesity and type 2 diabetes is growing at an alarming rate, including among pregnant women. Low-calorie sweeteners (LCSs) have increasingly been used as an alternative to sugar to deliver a sweet taste without the excessive caloric load. However, there is little evidence regarding their biological effects, particularly during development. Here, we used a mouse model of maternal LCS consumption to explore the impact of perinatal LCS exposure on the development of neural systems involved in metabolic regulation. We report that adult male, but not female, offspring from both aspartame- and rebaudioside A-exposed dams displayed increased adiposity and developed glucose intolerance. Moreover, maternal LCS consumption reorganized hypothalamic melanocortin circuits and disrupted parasympathetic innervation of pancreatic islets in male offspring. We then identified phenylacetylglycine (PAG) as a unique metabolite that was upregulated in the milk of LCS-fed dams and the serum of their pups. Furthermore, maternal PAG treatment recapitulated some of the key metabolic and neurodevelopmental abnormalities associated with maternal LCS consumption. Together, our data indicate that maternal LCS consumption has enduring consequences on the offspring's metabolism and neural development and that these effects are likely to be mediated through the gut microbial co-metabolite PAG.

Keywords: Imprinting; Melanocortin; Metabolism; Neuroendocrine regulation; Neuroscience.

Figure 1. Impact of LCS consumption during pregnancy and lactation on the dam’s metabolism.

(A) Experimental overview of the maternal LCS consumption mouse model. (B) Body weight curves, (C) average food intake, and (D) liquid intake in dams consuming a water (control), rebaudioside A, or aspartame solution during pregnancy and lactation (n = 5–7 dams per group). (E) Fat mass, (F) lean mass, and (G) glucose tolerance tests at the end of lactation in dams consuming a water (control), rebaudioside A, or aspartame solution (n = 4–6 dams per group). (H) Quantification of islet size and (I) percentage of β cell mass in the pancreas of dams consuming a water (control), rebaudioside A, or aspartame solution (n = 5–6 dams per group). (J) Serum insulin level at the end of lactation and (K) serum leptin level at P14 of dams consuming a water (control), rebaudioside A, or aspartame solution (n = 3–5 dams per group). Data are presented as mean ± SEM. Statistical significance between groups was determined by 2-way ANOVA followed by Bonferroni’s multiple-comparison test (B–D and G) or 1-way ANOVA followed by Tukey’s multiple-comparison test (E, F, and H–K). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001.

Figure 2. Maternal LCS consumption alters body composition and causes glucose intolerance in the male offspring.

(A) Litter size of dams consuming a water (control), rebaudioside A, or aspartame solution (n = 4–7 dams per group). (B) Body weight curves of male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 9–10 animals per group). (C) Fat and (D) lean mass of 14-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 6–10 per group). (E) Glucose tolerance tests and (F) areas under the curve of 11- to 12-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 7–9 animals per group). (G and H) Serum leptin and (I and J) insulin levels in (G and I) P14 and (H and J) 14-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 6–15 animals per group). (K) Food intake, (L) water intake, (M) locomotive activity, (N) respiratory exchange ratio (RER), and (O) energy expenditure in 14-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 6–13 animals per group). Data are presented as mean ± SEM. Statistical significance between groups was determined by 2-way ANOVA followed by Bonferroni’s multiple-comparison test (B, E, and N) or 1-way ANOVA followed by Tukey’s multiple-comparison test (A, C, D, F–M, and O). *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001.

Figure 3. Maternal LCS consumption reorganizes hypothalamic melanocortin circuits.

(A and B) Confocal images and quantification of the density of POMC- and AgRP-immunoreactive fibers innervating (A) the paraventricular nucleus of the hypothalamus (PVH) and (B) the dorsomedial nucleus (DMH) of 14-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 6–9 animals per group). The white dotted area illustrates the quantified area. Data are presented as mean ± SEM. Statistical significance between groups was determined by 1-way ANOVA followed by Tukey’s multiple-comparison test. *P ≤ 0.05, and **P ≤ 0.01. Scale bars: 50 μm. V3, third ventricle.

Figure 4. Maternal LCS consumption disrupts pancreatic parasympathetic innervation in the offspring.

(A and B) Confocal images and quantifications of the density of (A) VAChT-immunoreactive (shown in red) and (B) TH-immunoreactive fibers (red) in insulin⁺ (green) islets of 14-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 6–8 animals per group). The white dotted area illustrates the quantified area. (C) Quantification of islet size and (D) percentage of β cell mass in the pancreas of 14-week-old male mice born to dams consuming a water (control), rebaudioside A, or aspartame solution (n = 6–8 animals per group). Data are presented as mean ± SEM. Statistical significance between groups was determined by 1-way ANOVA followed by Tukey’s multiple-comparison test. *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Scale bars: 50 μm.

Figure 5. Maternal LCS consumption changes the metabolites in the adult offspring mice. (A

and B) Heatmap of significantly enriched (P ≤ 0.05; FDR uncorrected) putative metabolites in (A) the breast milk of dams consuming a water (control), rebaudioside A, or aspartame solution (n = 4 animals per group) and (B) the serum of their pups at P10 (n = 7–8 animals per group).

Figure 6. PAG is the only metabolite found in both the milk of LCS-fed dams and the serum of their pups.

(A) Venn diagrams showing metabolites that are significantly enriched in the milk of dams consuming either a rebaudioside A (RebA) or an aspartame (Asp) solution and the serum of their pups at P10. (B) Fold-changes in PAG levels in the milk of dams consuming a water (control), rebaudioside A, or aspartame solution and the serum of their pups at P10 (n = 4 dams and n = 8 pups per groups). Data are presented as mean ± SEM.

Figure 7. Maternal PAG treatment causes glucose intolerance and alters hypothalamic and autonomic circuits in the male offspring.

(A) Experimental overview of the maternal PAG treatment. (B) Plasma PAG levels 30 minutes after intraperitoneal injection of 36.5 mg/kg of PAG in C57BL/6J female mice (n = 4–5 per group). (C) Body weight curves, (D) perigonadal fat mass, and (E) glucose tolerance tests of dams treated with PAG (36.5 mg/kg) or vehicle (control) (n = 5–6 dams per group). (F) Body weight curves, (G) perigonadal fat mass, and (H) glucose tolerance tests of 14-week-old male mice born to PAG- or vehicle-treated dams (n = 12–21 animals per group). (I) Percentage of live cells in hypothalamic mHypoE-N43/5 cells treated with saline or 15, 70, 150, 300, or 400 μM of PAG (n = 5 independent cultures per condition). (J) Confocal images and quantification of TUJ1⁺ (neuron-specific class III β-tubulin) fibers derived from isolated ARH explants incubated with vehicle or PAG (400 μM) (n = 7 independent cultures per condition). (K) Photomicroscopic images and quantification of the density of POMC and AgRP fibers innervating the PVH of 14-week-old male mice born to dams treated with PAG or vehicle (n = 7–10 animals per group). (L) Confocal images and quantifications of VAChT-immunoreactive (shown in red) and TH-immunoreactive fibers’ (red) density in insulin⁺ (green) islets of 14-week-old male mice born to PAG- or vehicle-treated dams (n = 4–7 animals per group). Data are presented as mean ± SEM. Statistical significance between groups was determined by 2-way ANOVA followed by Bonferroni’s multiple-comparison test (C, E, F, H, and I) or unpaired 2-tailed Student’s t test (B, D, G, and J–L). *P ≤ 0.05, **P ≤ 0.01, and ****P ≤ 0.0001. Scale bars: 50 μm.