Metformin triggers a kidney GDF15-dependent area postrema axis to regulate food intake and body weight

Abstract

Metformin, the most widely prescribed medication for obesity-associated type 2 diabetes (T2D), lowers plasma glucose levels, food intake, and body weight in rodents and humans, but the mechanistic site(s) of action remain elusive. Metformin increases plasma growth/differentiation factor 15 (GDF15) levels to regulate energy balance, while GDF15 administration activates GDNF family receptor α-like (GFRAL) that is highly expressed in the area postrema (AP) and the nucleus of the solitary tract (NTS) of the hindbrain to lower food intake and body weight. However, the tissue-specific contribution of plasma GDF15 levels after metformin treatment is still under debate. Here, we found that metformin increased plasma GDF15 levels in high-fat (HF) fed male rats through the upregulation of GDF15 synthesis in the kidney. Importantly, the kidney-specific knockdown of GDF15 expression as well as the AP-specific knockdown of GFRAL expression negated the ability of metformin to lower food intake and body weight gain. Taken together, we unveil the kidney as a target of metformin to regulate energy homeostasis through a kidney GDF15-dependent AP axis.

Keywords: GDF15; area postrema; body weight; brain; food intake; kidney; liver; metformin; small intestine.

Figure 1.. The effect of USI, ileal, and IV infusion of metformin on food intake, body weight, and GDF15

(A and B) Cumulative food intake (A) and body weight gain (B) during refeeding in the rats after receiving USI saline (n = 11–19) or metformin (n = 8–16) infusion. (C) Plasma GDF15 levels in the rats at 6 h after USI saline (n = 8) or metformin (n = 8) infusion. (D) Gdf15 mRNA expression in the USI mucosal, ileal mucosal, liver, and kidney tissues in the rats at 6 h after USI saline (USI/liver n = 7, ileum/kidney n = 8) or metformin (USI/ileum/kidney n = 9, liver n = 7) infusion. (E and F) Cumulative food intake (E) and body weight gain (F) during refeeding in the rats after receiving ileal saline (n = 6–12) or metformin (n = 8–16) infusion. (G) Plasma GDF15 levels in the rats at 6 h after ileal saline (n = 8) or metformin (n = 8) infusion. (H) Gdf15 mRNA expression in the USI mucosal, ileal mucosal, liver, and kidney tissues in the rats at 6 h after ileal saline (USI/ileum/liver/kidney n = 8) or metformin (USI/ileum/liver/kidney n = 8) infusion. (I and J) Cumulative food intake (I) and body weight gain (J) during refeeding in the rats after receiving IV saline (n = 7–20) or metformin (n = 10–26) infusion. (K) Plasma GDF15 levels in the rats at 6 h after IV saline (n = 8) or metformin (n = 8) infusion. (L) Gdf15 mRNA expression in the USI mucosal, ileal mucosal, liver, and kidney tissues in the rats at 6 h after IV saline (USI/ileum/kidney n = 8, liver n = 6) or metformin (USI/ileum/liver/kidney n = 8) infusion. (M) Plasma GDF15 levels in the rats at 1 h after IV saline (n = 6) or metformin (n = 6) infusion. (N) Gdf15 mRNA expression in the USI mucosal, ileal mucosal, liver, and kidney tissues in the rats at 1 h after IV saline (USI/ileum/liver n = 6, kidney n = 5) or metformin (USI n = 8, ileum/kidney n = 6, liver n = 7) infusion. (A)—(N) *p < 0.05, **p < 0.01 vs. relative saline group calculated by unpaired Student’s t tests. Sal, saline; Met, metformin; Ile, ileum. Data are presented as mean ± SEM. See also Figure S1.

Figure 2.. The effect of USI or kidney-specific knockdown of GDF15 on metformin action

(A) Gdf15 mRNA expression in the USI mucosal tissues taken at 6 h after USI saline/metformin infusion in the USI shMM (saline n = 6, metformin n = 7) or shGDF15 (saline n = 6, metformin n = 7) lentiviral infected rats. (B and C) Cumulative food intake during refeeding in the USI shMM (B)/shGDF15 (C) lentiviral infected rats after receiving USI saline (shMM/shGDF15 n = 7) or metformin (shMM/shGDF15 n = 8) infusion. (D) Body weight gain at 12 h during refeeding in the USI shMM/shGDF15 lentiviral infected rats after receiving USI saline (shMM/shGDF15 n = 7) or metformin (shMM/shGDF15 n = 8) infusion. (E) Plasma GDF15 levels in the USI shMM/shGDF15 lentiviral infected rats at 6 h after USI saline (shMM/shGDF15 n = 7) or metformin (shMM/shGDF15 n = 8) infusion. (F) Gdf15 mRNA expression in the kidney tissues taken at 6 h after IV metformin infusion in the kidney shMM (n = 6) or shGDF15 (n = 7) lentiviral infected rats. (G and H) Cumulative food intake during refeeding in the kidney shMM (G)/shGDF15 (H) lentiviral infected rats after receiving IV saline (shMM n = 9, shGDF15 n = 12) or metformin (shMM n = 10, shGDF15 n = 11) infusion. (I) Body weight gain at 6 h during refeeding in the kidney shMM/shGDF15 lentiviral infected rats after receiving IV saline (shMM n = 9, shGDF15 n = 12) or metformin (shMM n = 10, shGDF15 n = 11) infusion. (J) Plasma GDF15 levels in the kidney shMM/shGDF15 lentiviral infected rats at 1 h after IV saline (shMM/shGDF15 n = 6) or metformin (shMM/shGDF15 n = 6) infusion. (K) Plasma GDF15 levels in the kidney shMM/shGDF15 lentiviral infected rats at 6 h after IV saline (shMM/shGDF15 n = 6) or metformin (shMM/shGDF15 n = 6) infusion. (A), (D), (E), and (I)—(K) *p < 0.05, **p < 0.01 vs. saline or indicated groups, and #p < 0.05, ##p < 0.01 vs. shMM + relative saline/metformin group calculated by one-way ANOVA with Tukey post hoc test; (B), (C), (G), and (H) *p < 0.05, **p < 0.01 vs. saline group calculated by unpaired Student’s t tests; (F) *p < 0.05 vs. shMM group calculated by unpaired Student’s t tests. Sal, saline; Met, metformin; Kid, kidney. Data are presented as mean ± SEM. See also Figure S2.

Figure 3.. Kidney GDF15 contribution to long-term metformin treatment and its expression in humans

(A and B) Daily cumulative food intake in the kidney shMM (A)/shGDF15 (B) lentiviral infected rats receiving daily intragastric saline (shMM/shGDF15 n = 9) or metformin (shMM n = 10, shGDF15 n = 8) administration for 11 days. (C) Percent change (vs. relative saline groups) of daily cumulative food intake in the kidney shMM (n = 10) or shGDF15 (n = 8) lentiviral infected rats receiving daily intragastric metformin administration for 11 days. (D and E) Daily cumulative body weight gain in the kidney shMM (D)/shGDF15 (E) lentiviral infected rats receiving daily intragastric saline (shMM/shGDF15 n = 9) or metformin (shMM n = 10, shGDF15 n = 8) administration for 11 days. (F) Percent change (vs. relative saline groups) of daily cumulative body weight gain in the kidney shMM (n = 10) or shGDF15 (n = 8) lentiviral infected rats receiving daily intragastric metformin administration for 11 days. (G) Plasma GDF15 levels in the kidney shMM/shGDF15 lentiviral infected rats at 6 h after the last dose of saline (shMM/shGDF15 n = 9) or metformin (shMM/shGDF15 n = 8) administration on day 11. (H) Gdf15 mRNA expression in the kidney tissues taken at 6 h after the last dose of saline (shMM/shGDF15 n = 8) or metformin (shMM n = 7, shGDF15 n = 8) administration on day 11 in the kidney shMM/shGDF15 lentiviral infected rats. (I) Gdf15 mRNA expression in the liver tissues taken at 6 h after the last dose of saline (shMM/shGDF15 n = 8) or metformin (shMM/shGDF15 n = 8) administration on day 11 in the kidney shMM/shGDF15 lentiviral infected rats. (J–L) Metformin concentrations in the plasma (J), liver (K), and kidney (L) tissues taken at 6 h after the last dose of saline (shMM n = 9) or metformin (shMM n = 10, shGDF15 n = 8) administration on day 11 in the kidney shMM/shGDF15 lentiviral infected rats. (M) Gdf15 mRNA expression in the tubulo-interstitial (n = 337) and glomerular (n = 257) tissue of kidney biopsies in humans. (N and O) Correlations between tubulo-interstitial Gdf15 mRNA expression and eGFR (N, n = 331)/random serum glucose (O, n = 282) in humans. (A) and (D) *p < 0.05, **p < 0.01 vs. saline group calculated by unpaired Student’s t tests; (C) and (F) *p < 0.05, **p < 0.01 vs. shMM group calculated by unpaired Student’s t tests; (G) and (H) **p < 0.01 vs. saline groups, and #p < 0.05 vs. shMM + metformin group calculated by one-way ANOVA with Tukey post hoc test; (M) p values calculated by unpaired Student’s t tests; (N) and (O) p and Rho values calculated by Spearman’s rank correlation coefficient. Sal, saline; Met, metformin; Kid, kidney; Gastra, intragastric. Data are presented as mean ± SEM, box plot, or scatter plot. See also Figure S3 and Table S1.

Figure 4.. The sufficiency and necessity of the GDF15-GFRAL axis in the AP

(A and B) Cumulative food intake (A) and body weight gain (B) during refeeding in the rats after receiving AP saline (n = 10) or GDF15 (n = 10) infusion. (C and D) Cumulative food intake (C) and body weight gain (D) during refeeding in the DVC shMM/shGFRAL lentiviral infected rats after receiving AP saline (shGFRAL n = 6) or GDF15 (shMM n = 8, shGFRAL n = 6) infusion. (E and F) GFRAL mRNA expression in the AP (E) or DVC without AP (F) tissues taken at 24 h of refeeding in the AP shMM (AP/DVC without AP n = 8)/shGFRAL (AP/DVC without AP n = 8) lentiviral infected rats. (G and H) Cumulative food intake during refeeding in the AP shMM (G)/shGFRAL (H) lentiviral infected rats after receiving IV saline (shMM n = 10, shGFRAL n = 12) or metformin (shMM/shGFRAL n = 11) infusion. (I) Body weight gain at 6 h during refeeding in the AP shMM/shGFRAL lentiviral infected rats after receiving IV saline (shMM n = 10, shGFRAL n = 12) or metformin (shMM/shGFRAL n = 11) infusion. (J and K) Cumulative food intake during refeeding in the AP shMM (J)/shGFRAL (K) lentiviral infected rats after receiving USI saline (shMM n = 10, shGFRAL n = 11) or metformin (shMM n = 10, shGFRAL n = 11) infusion. (L) Body weight gain at 12 h during refeeding in the AP shMM/shGFRAL lentiviral infected rats after receiving USI saline (shMM n = 10, shGFRAL n = 11) or metformin (shMM n = 10, shGFRAL n = 11) infusion. (A), (B), (G), (H), (J), and (K) *p < 0.05, **p < 0.01 vs. saline group calculated by unpaired Student’s t tests; (C) and (D) *p < 0.05, **p < 0.01 vs. AP shGFRAL + AP saline/GDF15 groups calculated by two-way ANOVA with Tukey post hoc test; (E) *p < 0.05 vs. shMM group calculated by unpaired Student’s t tests; (I) and (L) *p < 0.05, **p < 0.01 vs. saline groups, and #p < 0.05 vs. shMM + metformin group calculated by one-way ANOVA with Tukey post hoc test. Sal, saline; Met, metformin. Data are presented as mean ± SEM. See also Figure S4.