GDF15 Induces Anorexia through Nausea and Emesis

Abstract

Growth differentiation factor 15 (GDF15) is a cytokine that reduces food intake through activation of hindbrain GFRAL-RET receptors and has become a keen target of interest for anti-obesity therapies. Elevated endogenous GDF15 is associated with energy balance disturbances, cancer progression, chemotherapy-induced anorexia, and morning sickness. We hypothesized that GDF15 causes emesis and that its anorectic effects are related to this function. Here, we examined feeding and emesis and/or emetic-like behaviors in three different mammalian laboratory species to help elucidate the role of GDF15 in these behaviors. Data show that GDF15 causes emesis in Suncus murinus (musk shrews) and induces behaviors indicative of nausea/malaise (e.g., anorexia and pica) in non-emetic species, including mice and lean or obese rats. We also present data in mice suggesting that GDF15 contributes to chemotherapy-induced malaise. Together, these results indicate that GDF15 triggers anorexia through the induction of nausea and/or by engaging emetic neurocircuitry.

Keywords: GDF15; GFRAL; MIC-1; Suncus murinus; anorexia; chemotherapy; emesis; malaise; nausea; obesity.

Figure 1.. Cisplatin Induces Higher Circulating GDF15 Levels and Activates GFRAL Positive Neurons in the AP/NTS of the Mouse

(A) A single dose of cisplatin (10 mg/kg) induced a chronic anorectic response (n = 4–5 per group). (B) Body weight between cisplatin-treated and pair-fed controls did not differ (n = 4–5 per group). (C) Plasma GDF15 levels were higher at 2, 6, and 24 h after cisplatin injection and slowly returned to baseline 7 days after (n = 4–8 per group). (D) The magnitude of anorexia across time positively correlated with circulating GDF15 levels (n = 15). (E) Representative sections of the AP/NTS showing the distribution of GFRAL neurons and the colocalization with of the marker of neuronal activation c-Fos 24 h after cisplatin or saline injection. (F) Quantification of cisplatin-induced c-Fos expression in the AP/NTS 24 h upon injection (n = 4–5 per group). (G) Number of GFRAL-positive neurons in the AP/NTS that co-express c-Fos 24 h after cisplatin administration (n = 4–5 per group). (H) Representative sections of the AP/NTS depicting GFRAL neurons and c-Fos 3 days after cisplatin or saline injection. (I) Quantification of c-Fos expression in the AP/NTS 3 days after drug administration (n = 5 per group). (J) Number of GFRAL-positive neurons in the AP/NTS that co-express c-Fos 3 days upon treatment (n = 5 per group). (K) Representative sections of the AP/NTS showing GFRAL neurons and c-Fos 7 days after treatment. (L) Quantification of c-Fos expression in the AP/NTS 7 days after drug administration (n = 5 per group). (M) Number of GFRAL-positive neurons in the AP/NTS that co-express c-Fos 7 days after treatment (n = 5 per group). All data but in (D) are expressed as mean ± SEM. Data in (C) analyzed using two-way ANOVA followed by Tukey’s post-hoc test. Means with different letters are significantly different (p < 0.05). Quantification data in (F, G, I, J, L, and M) analyzed with the Student’s t-test; *p < 0.05, ***p < 0.0001. Arrows represent c-Fos/ GFRAL neurons. Scale bar 100 µm. Inset in (E) is a high-power (60x) photomicrograph showing a GFRAL-positive neuron co-expressing c-Fos in the AP. Scale bar 10 mm.

Figure 2.. Hepatic, but Not Renal, Duodenal, nor Splenic GDF15 mRNA Expression Is Upregulated Following Cisplatin

(A) Hepatic GDF15 mRNA was higher at day 1 after cisplatin injection compared to controls. No significant differences in GDF15 mRNA expression occurred at day 3 and day 7. (B) Hepatic TNF-a mRNA was higher at day 1 after cisplatin injection compared to controls. No significant differences occurred at later timepoints. (C) In cisplatin-treated animals, COX-2 mRNA expression in the liver was higher at 1 day, lower at 3 days, and unaltered at 7 days compared to the relative PF groups. (D) GDF15 mRNA levels in the kidney, duodenum, and spleen were unaltered at day 1 upon cisplatin administration compared to controls. (E) TNF-a mRNA expression levels in the kidney, duodenum, and spleen were not affected by cisplatin administration at day 1. (F) COX-2 mRNA transcripts were lower in the kidney and duodenum at day 1, while no change occurred in the spleen compared to their relative controls. All data expressed as mean ± SEM. n = 4–5 per group. Data in (A–C) expressed as mean ± SEM and analyzed using two-way ANOVA followed by Bonferroni’s post-hoc test; *p < 0.05. Data in (D–F) analyzed with the Student’s t test; ***p < 0.0001.

Figure 3.. Central GDF15 Delivery Induces CFA, Anorexia, Body Weight Loss, and Pica in Rats

(A) The preference for the paired solution was significantly reduced after centrally delivered GDF15 (30 pmol i.c.v., n = 7) compared to negative controls (DMSO, 1 µL i.c.v., n = 4). LiCl (0.15 M, i.p., n = 6) was used as a positive control. (B) GDF15 (30 pmol i.c.v., n = 22) significantly reduced 24 h food intake but did not cause anorexia at 3 or 6 h post-delivery compared to the control group. (C) GDF15 administration triggered kaolin consumption at all time points and preceded the onset of the anorectic response. (D) GDF15-induced anorexia was paralleled by significant body weight loss at 24 h post-injection compared to controls. (E) GDF15 delivered at 10, 30, 60 pmol into the 4^th ventricle induced anorexia compared to vehicle-treated animals (n= 10 per group). (F) The same doses of GDF15 triggered kaolin consumption (i.e. pica) (n = 10 per group). (G) In a separate cohort of rats, GDF15 was infused at 0, 1, 3, 10 pmol. While 1 pmol had no effect, all other tested doses induced anorexia (n = 10 per group). (H) Similarly, all but the 1 pmol dose of GDF15 triggered pica (n = 10 per group). In general, pica preceded the onset of the hypophagic response across all the anorectic doses tested. All data expressed as mean ± SEM. Data in (B and C) were analyzed with a repeated-measurements two-way ANOVA followed by the Bonferroni’s post hoc test. Data in (D) were analyzed with the Student’s t test. **p < 0.01, ***p < 0.001. Data in (E–H) were analyzed with repeated-measure two-way ANOVA followed by Tukey’s post hoc test. Means with different letters are significantly different from each other (p < 0.05).

Figure 4.. Systemic, but Not Hindbrain, Pharmacological Inhibition of 5-HT3R Signaling Attenuates GDF15-Induced Anorexia and Body Weight Loss

(A and B) Systemic pre-treatment of the selective 5-HT3R antagonist Ondansetron (Ond, 1 mg/kg i.p., n = 13) attenuated anorexia (A), but failed to reduce kaolin consumption (B). (C) Systemic Ond was also effective in attenuating GDF15-induced body weight loss. (D–F) Ond i.c.v. pre-treatment (25 mg, n = 13) infused into the 4^th ventricle did not prevent GDF15-induced anorexia (D), pica (E), and body weight loss (F). Data expressed as mean ± SEM and analyzed with a 2 × 2 repeated-measurements two-way ANOVA followed by the Tukey’s post hoc test. Means with different letters are significantly different (p < 0.05).

Figure 5.. Central and Systemic Delivery of GDF15 Produces Anorexia, Pica, and Body Weight Loss in High-Fat/High-Sucrose Diet-Induced Obese Rats

(A–C) Central administration of GDF15 (10 and 30 pmol into the 4^th ventricle) dose-dependently induced anorexia (A), kaolin consumption (B), and body weight loss (C) in obese rats (n = 14) compared to vehicle treated animals. (D–F) Consistent with the central route of delivery, systemic administration of GDF15 (20 and 100 µg/kg i.p.) to obese rats (n = 14) induced anorexia (D), kaolin intake (E), and body weight loss (F) in a dose-dependent fashion. All data expressed as mean ± SEM. Data in (B and C) were analyzed with a repeated-measurements two-way ANOVA followed by the Bonferroni’s post hoc test. Means with different letters are significantly different from each other (p < 0.05).

Figure 6.. GDF15 Induces Severe Emesis and c-Fos Expression in the NTS but Mild Anorexia and Body Weight Loss in Shrews

(A) In seven of the eight animals tested, 1mg/kg of GDF15 induced robust emetic responses that were not observed after 0.1 mg/kg GDF15 or saline injections. (B) Graphical representation of latency to the first emetic episode of GDF15-treated animals that exhibited emesis. (C) Heatmap showing latency, number of emetic episodes per minute for each animal across time. (D) Representative immunofluorescent images showing c-Fos-positive neurons across in the dorsal vagal complex (250 µm rostral to the obex), 3 h after 1 mg/kg GDF15 (n = 3), 0.1 mg/kg GDF15 (n = 3), or vehicle (n = 5) i.p. injection. (E) Systemic GDF15 administration at 1 mg/kg lead to a significantly higher number of c-Fos immunoreactive cells in the NTS of shrews 3 h after injection but not in the AP or DMV. (F) GDF15 administered at 1 mg/kg only reduced food intake at 48 h while the lower dose of GDF15 had a minimal, yet significant, hypophagic effect at 4 h relative to controls. No significant differences were noticed at other time points. (G) In line with the food intake data, the lower dose of GDF15 induced body weight loss at 24 and 48 h compared to vehicle. The highest dose tested induced significant body weight loss at 48 h. Data in (A) were analyzed with repeated-measurements one-way ANOVA followed by Tukey’s post hoc test. Data in (E) were analyzed with one-way ANOVA followed by Tukey’s post hoc test. Data in (G and F) were analyzed with repeated-measurements two-way ANOVA followed by Tukey’s post hoc test. All data expressed as mean ± SEM. Means with different letters are significantly different (p < 0.05). Scale bar, 100 µm.