GDF15 Provides an Endocrine Signal of Nutritional Stress in Mice and Humans

Abstract

GDF15 is an established biomarker of cellular stress. The fact that it signals via a specific hindbrain receptor, GFRAL, and that mice lacking GDF15 manifest diet-induced obesity suggest that GDF15 may play a physiological role in energy balance. We performed experiments in humans, mice, and cells to determine if and how nutritional perturbations modify GDF15 expression. Circulating GDF15 levels manifest very modest changes in response to moderate caloric surpluses or deficits in mice or humans, differentiating it from classical intestinally derived satiety hormones and leptin. However, GDF15 levels do increase following sustained high-fat feeding or dietary amino acid imbalance in mice. We demonstrate that GDF15 expression is regulated by the integrated stress response and is induced in selected tissues in mice in these settings. Finally, we show that pharmacological GDF15 administration to mice can trigger conditioned taste aversion, suggesting that GDF15 may induce an aversive response to nutritional stress.

Keywords: GDF15; GFRAL; conditioned taste aversion; integrated stress response; overnutrion.

Graphical abstract

Figure 1

GDF15 Levels in Response to a Meal or Imposed Caloric Deficit in Mice and Humans (A–D) Human Study 1 (HS1): (A) plasma glucose, (B) insulin, (C) GLP-1, and (D) GDF15 circulating levels in six healthy volunteers given an oral mixed macronutrient liquid meal following an overnight fast. Blood samples were taken over the 180 min duration of the study. Data are expressed as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001 compared to time 0 min by one-way ANOVA with Bonferroni post-test. (E–G) Mouse Study 1 (MS1): (E) body weight of mice before and after a 24 h fasting challenge (n = 5 mice). (F) Leptin and (G) GDF15 serum concentrations in 11- to 12-week-old male mice either in the fed state or after a 24 h fast. Data are expressed as mean ± SEM (n = 5 mice per group). ^∗∗∗p < 0.001 by two-tailed Student’s t test. Note all fasted leptin values were under the detection limit (0.033 ng/mL). (H–J) Human Study 3 (HS3): (H) body weight, (I) leptin, and (J) GDF15 levels in a cohort of overweight and obese participants subjected to caloric restriction (∼1,000 kcal/day) for a period of 28 days. Data are from 33 participants, expressed as mean ± SEM. ^∗p < 0.05, ^∗∗∗∗p < 0.0001 by a one-way ANOVA with Bonferroni multiple comparison post-test (for body weight and leptin). (K–M) Human Study 4 (HS4): (K) leptin, (L) β-hydroxybutyrate, and (M) GDF15 levels in human volunteers subjected to a 7-day fast (0 kcal per day). Data are from 13 participants, expressed as mean ± SEM and analyzed by a one-way ANOVA. In the case of GDF15, values are expressed as median (interquartile range). ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by Kruskal-Wallis test. See also Figures S1 and S4.

Figure 2

GDF15 Is Upregulated by Long-Term High-Fat Feeding in Mice (A–F) Mouse Study 3 (MS3): C57BL/6J male mice (aged 9 weeks) were fed a chow (CD) or high-fat diet (HFD) for 16 weeks. (A) Body weight was recorded weekly (CD, n = 7; HFD, n = 8), while (B) fat mass and (C) insulin, (D) leptin, (E) glucose, and (F) GDF15 concentrations were determined at 0, 4, 8, 12 (CD, n = 9–11; HFD, n = 10–12), and 16 weeks (CD, n = 7; HFD, n = 8) (all after a 4 h fast). Data are expressed as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by two-way ANOVA with Bonferroni multiple comparison post-test. The red asterisks in (D) denote time points at which some (1 out of 12 at 12 weeks and 3 out of 8 at 16 weeks) leptin values were above the assay detection limit (>100 ng/mL) and thus were set at 100 ng/mL. (G) GDF15 mRNA expression in subcutaneous (SAT), epididymal (EAT), and brown (BAT) adipose tissue, liver, soleus muscle, and kidney isolated from C57BL/6J male mice fed a CD or HFD for 18 weeks (n = 6–8 mice/group). mRNA is presented as fold expression (mean ± SEM) relative to the chow-fed state from muscle (set at 1) and normalized to the geometric mean of B2M/36b4 gene expression. ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 by two-tailed Student’s t test. See also Figures S2 and S4.

Figure 3

GDF15 Expression Is Regulated by the Cellular ISR Pathway (A and C) GDF15 mRNA expression (A) and immunoblot analysis (C) of ISR components in wild-type (WT) mouse embryonic fibroblasts (MEFs) treated with vehicle control (Con), cobalt chloride (CoCl2, 625 μM), thapsigargin (Tg, 1 μM), tunicamycin (Tn 5 μg/mL), or L-Histidinol (His, 1 mM) for 6 h. (B) GDF15 mRNA expression in human cell lines (HeLA, HuH7, and A549) treated with Tn (5 μg/mL) for 6 h. (D) GDF15 mRNA expression in WT MEFs pre-treated for 1 h either with the PERK inhibitor GSK2606414 (GSK, 200 nM) or eIF2α inhibitor ISRIB (ISR, 100 nM), then co-treated with Tn (5 μg/mL) for a further 6 h. (E–G) GDF15 mRNA expression (E) in EIF2α Ser51 (SS) or phospho mutant (AA) MEFs or (F) in ATF4 wild-type (WT) or ATF4 knockout (KO) MEFs and (G) in control siRNA and CHOP siRNA transfected WT MEFs treated with Tn (5 μg/mL) for 6 h. (H) Diagram outlining pathway by which GDF15 and FGF21 expression is regulated by TN. mRNA expression is presented as fold expression relative to its respective control treatment for each cell type (set at 1) or TN-treated samples (set as 100) with normalization to HPRT gene expression in MEFs and GAPDH in human cells. Data are expressed as mean ± SD from at least three independent experiments. ^∗∗∗p < 0.001 versus control (con) for (A) and (B), and versus TN stimulated for (D)–(G) by two-tailed Student’s t test. Blots shown are representative of three independent experiments with Calnexin used as a loading control. See also Figures S3 and S4.

Figure 4

GDF15 Is Upregulated in Response to a Lysine-Deficient Diet and Induces Conditioned Taste Aversion in Mice (A–D) Mouse Study 4 (MS4): (A) GDF15 serum concentrations and (B) ATF4, (C) CHOP, and (D) GDF15 mRNA expression in livers of 11- to 12-week-old female mice that were fasted overnight and then fed a control chow (Con) or lysine-deficient diet (−Lys) for 4 h. A blood sample was withdrawn at 1 h following the beginning of the meal. Serum and mRNA (4 h time point only) data are expressed as mean ± SEM (n = 6 mice per group) with mRNA normalized to B2M gene expression. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001 by one-way ANOVA. (E) Mouse Study 5 (MS5): Circulating plasma GDF15 concentrations after a single dose of recombinant GDF15 in mice; dose response (n = 3/ group). (F and G) Cumulative food intake measured between 1 and 4 h post-GDF15 dose expressed as total grams (F) or percent (%) of vehicle control (G) (n = 7–8/group). Data are presented as mean ± SEM. ^∗∗∗∗p < 0.0001 versus vehicle by one-way ANOVA with Bonferroni multiple comparison post-test. (H) Saccharin and water consumption during conditioned taste aversion test during GDF15 treatment (n = 8–16/group). Data are presented as mean ± SEM and analyzed using a two-way ANOVA with Bonferroni multiple comparison post-test to compare proportion of saccharin water and water consumption between groups of GDF15 or LiCl treatment to vehicle. ^∗∗∗∗(saccharin) or ^####(water) p < 0.0001. See also Figure S4.