The gastrointestinal tract is a major source of the acute metformin-stimulated rise in GDF15

Abstract

The hormone GDF15 is secreted in response to cellular stressors. Metformin elevates circulating levels of GDF15, an action important for the drug's beneficial effects on body weight. Metformin can also inhibit mammalian respiratory complex I, leading to decreases in ATP:AMP ratio, activation of AMP Kinase (AMPK), and increased GDF15 production. We undertook studies using a range of mice with tissue-specific loss of Gdf15 (namely gut, liver and global deletion) to determine the relative contributions of two classical metformin target tissues, the gut and liver, to the elevation of GDF15 seen with metformin. In addition, we performed comparative studies with another pharmacological agent, the AMP kinase pan-activator, MK-8722. Deletion of Gdf15 from the intestinal epithelium significantly reduced the circulating GDF15 response to oral metformin, whereas deletion of Gdf15 from the liver had no effect. In contrast, deletion of Gdf15 from the liver, but not the gut, markedly reduced circulating GDF15 responses to MK-8722. Further, our data show that, while GDF15 restricts high-fat diet-induced weight gain, the intestinal production of GDF15 is not necessary for this effect. These findings add to the body of evidence implicating the intestinal epithelium in key aspects of the pharmacology of metformin action.

Figure 1

Gut-targeted Gdf15 ablation results in significant reductions of acute metformin-stimulated circulating GDF15, a response altered by antecedent diet. (A) Serum GDF15 levels (mean ± SEM) in 12-week-old wild-type and Gdf15-gut-KO mice fed either a standard chow (Chow), 60% high-fat (HF), or high-fat high-sucrose (HFHS) diet for 4 weeks followed by a single oral dose of water (sham) or 600 mg/kg metformin, n = 9/group, and ns = non-significant, ***p < 0.001, and ****p < 0.0001 as determined by three-way ANOVA. (B–G) Gdf15 mRNA expression (normalized to expression levels of actin B) in the (B) liver, (C) kidney, (D) colon, (E) proximal small intestine, (F) middle small intestine, and (G) distal small intestine from 60% high-fat diet-fed wild-type and Gdf15-gut-KO mice 6 h after receiving a single oral dose of water (Sham) or 600 mg/kg metformin (Met). n = 9/group, mean ± SEM, and *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 as determined by two-way ANOVA.

Figure 2

Loss of liver Gdf15 does not impact the acute metformin-stimulated rise in circulatory GDF15. (A) Body weight of 12-week-old wild-type (n = 17) and Gdf15-liver-KO (n = 10) mice on a 60% high-fat diet for 4 weeks beginning at 8 weeks of age, mean ± SEM. (B) Serum GDF15 levels (mean ± SEM) in 12-week-old wild-type and Gdf15-liver-KO mice fed a 60% high-fat diet for 4 weeks followed by a single oral dose of water (sham) or 600 mg/kg metformin, n = 9/group, and ns = non-significant, ***p < 0.001, and ****p < 0.0001 as determined by two-way ANOVA. (C–H) Gdf15 mRNA expression (normalized to expression levels of actin B) in the (C) liver, (D) kidney, (E) colon, (F) proximal small intestine, (G) middle small intestine, and (H) distal small intestine from 60% high-fat diet-fed wild-type and Gdf15-liver-KO mice 6 h after receiving a single oral dose of water (Sham) or 600 mg/kg metformin (Met). n = 9/group, mean ± SEM, and *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 as determined by two-way ANOVA.

Figure 3

The liver is a major contributor to the increase in circulating GDF15 in response to oral administration of an AMPK activator. (A) Serum GDF15 levels (mean ± SEM) in 20-week-old wild-type and Gdf15-liver-KO mice fed a 60% high-fat diet for 4 weeks followed by a single oral dose of water (sham) or 30 mg/kg MK-8722, n = 8/group, mean ± SEM, and ns = non-significant and ****p < 0.0001 as determined by two-way ANOVA. (B–G) Gdf15 mRNA expression (normalized to expression levels of actin B) in the (B) liver, (C) kidney, (D) colon, (E) proximal small intestine, (F) middle small intestine, and (G) distal small intestine from high-fat diet-fed wild-type and Gdf15-liver-KO mice 4 h after receiving a single oral dose of water (Sham) or 30 mg/kg MK-8722. n = 8/group, mean ± SEM, and ns = non-significant, *p < 0.05, and ****p < 0.0001 as determined by two-way ANOVA.

Figure 4

The gut does not contribute significantly to the protective effect of GDF15 on weight gain in response to a high fat diet. (A) Body weight of 23-week-old wild-type (n = 18), Gdf15-gut-KO (n = 15), and Gdf15-global-KO (n = 16) mice fed a 60% high-fat diet for 15 weeks beginning at 8 weeks of age, mean ± SEM. Inset, final body weight. (B) Percent body weight change of 23-week-old wild-type (n = 18), Gdf15-gut-KO (n = 15), and Gdf15-global-KO (n = 16) mice fed a 60% high-fat diet for 15 weeks beginning at 8 weeks of age, mean ± SEM. From 9 to 15 weeks on diet, global-KO body weight relative to wild-type mice p-value listed as determined by two-way ANOVA with multiple comparison adjustment by Tukey’s test. (C) Concentration of serum GDF15 (pg/mL) in 23-week-old wild-type (n = 18), Gdf15-gut-KO (n = 15), and Gdf15-global-KO (n = 16) mice fed a 60% high-fat diet for 15 weeks beginning at 8 weeks of age, mean ± SEM. ns = nonsignificant, ***p < 0.001, and ****p < 0.0001 as determined by two-way ANOVA with multiple comparison adjustment by Tukey’s test. (D) Concentration of serum FGF21 (pg/mL) in wild-type (n = 18), Gdf15-gut-KO (n = 15), and Gdf15-global-KO (n = 16) mice fed a 60% high-fat diet for 15 weeks beginning at 8 weeks of age, mean ± SEM. *p < 0.05 as determined by two-way ANOVA with multiple comparison adjustment by Tukey’s test.