FGF21 is increased by inflammatory stimuli and protects leptin-deficient ob/ob mice from the toxicity of sepsis

Abstract

The acute phase response (APR) produces marked alterations in lipid and carbohydrate metabolism including decreasing plasma ketone levels. Fibroblast growth factor 21 (FGF21) is a recently discovered hormone that regulates lipid and glucose metabolism and stimulates ketogenesis. Here we demonstrate that lipopolysaccharide (LPS), zymosan, and turpentine, which induce the APR, increase serum FGF21 levels 2-fold. Although LPS, zymosan, and turpentine decrease the hepatic expression of FGF21, they increase FGF21 expression in adipose tissue and muscle, suggesting that extrahepatic tissues account for the increase in serum FGF21. After LPS administration, the characteristic decrease in plasma ketone levels is accentuated in FGF21-/- mice, but this is not due to differences in expression of carnitine palmitoyltransferase 1α or hydroxymethyglutaryl-CoA synthase 2 in liver, because LPS induces similar decreases in the expression of these genes in FGF21-/- and control mice. However, in FGF21-/- mice, the ability of LPS to increase plasma free fatty acid levels is blunted. This failure to increase plasma free fatty acid could contribute to the accentuated decrease in plasma ketone levels because the transport of fatty acids from adipose tissue to liver provides the substrate for ketogenesis. Treatment with exogenous FGF21 reduced the number of animals that die and the rapidity of death after LPS administration in leptin-deficient ob/ob mice and to a lesser extent in control mice. FGF21 also protected from the toxic effects of cecal ligation and puncture-induced sepsis. Thus, FGF21 is a positive APR protein that protects animals from the toxic effects of LPS and sepsis.

Fig. 1.

Effect of APR stimuli on serum FGF21 levels and mRNA levels in mouse liver. Mice were injected with saline, LPS (5 mg/kg, ip), zymosan A (80 mg/kg, ip), or turpentine (100 μl/mouse, sc), and the animals were euthanized at 16 h. A, Serum FGF21 levels were measured by ELISA as described in Materials and Methods. B, Total RNA was isolated from liver tissue, cDNA was synthesized with reverse transcriptase, and quantitative real-time PCR was performed as described in Materials and Methods. FGF21 mRNA levels were measured. C, HMGCS2 mRNA levels were measured. Quantitative PCR data were normalized to 36B4 mRNA for all experiments. Data (mean ± se, n = 4) are expressed as a percentage of controls administered saline. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig. 2.

Effect of LPS on FGF21 mRNA in mouse liver. Mice were injected with saline or LPS at indicated doses and animals were euthanized. Total RNA was isolated from liver tissue, cDNA was synthesized with reverse transcriptase, and quantitative real-time PCR was performed as described in Materials and Methods. A, Animals were euthanized 4 h after LPS administration and FGF21 mRNA levels were measured. B, Animals were euthanized 16 h after LPS administration and FGF21 mRNA levels were measured. Quantitative PCR data were normalized to 36B4 mRNA for all experiments. Data (mean ± se, n = 4) are expressed as a percentage of controls administered saline. *, P < 0.05; ***, P < 0.001.

Fig. 3.

Effect of LPS on serum FGF21 and mRNA levels in mouse liver. Wild-type (WT) and PPARα-deficient mice were injected with either saline [con (control)] or LPS (5 mg/kg), and the animals were euthanized at 16 h after injection. Total RNA was isolated from liver tissue, cDNA was synthesized with reverse transcriptase, and quantitative real-time PCR was performed as described in Materials and Methods. A, FGF21 mRNA levels were measured. B, HMGCS2 mRNA levels were measured. Quantitative PCR data were normalized to 36B4 mRNA for all experiments. Quantitative PCR data (mean ± se, n = 4) are expressed as a percentage of controls administered saline. C, Serum FGF21 levels were measured by ELISA as described in Materials and Methods. Data (mean ± se, n = 5). **, P < 0.01; ***, P < 0.001 vs. control.

Fig. 4.

Effect of APR stimuli on mRNA levels in mouse adipose tissue. Mice were injected with saline, LPS (5 mg/kg, ip), zymosan A (80 mg/kg, ip), or turpentine (100 μl/mouse, sc), and the animals were euthanized at 16 h. Total RNA was isolated from adipose tissue, cDNA was synthesized with reverse transcriptase, and quantitative real-time PCR was performed as described in Materials and Methods. A, FGF21 mRNA levels were measured. B, GLUT1 mRNA levels were measured. Quantitative PCR data were normalized to 36B4 mRNA for all experiments. Data (mean ± se, n = 4) are expressed as a percentage of controls administered saline. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig. 5.

Effect of APR stimuli on mRNA levels in mouse muscle tissue. Mice were injected with saline, LPS (5 mg/kg, ip), zymosan A (80 mg/kg, ip), or turpentine (100 μl/mouse, sc), and the animals were euthanized at 16 h. Total RNA was isolated from muscle tissue, cDNA was synthesized with reverse transcriptase, and quantitative real-time PCR was performed as described in Materials and Methods. A, FGF21 mRNA levels were measured. B, GLUT1 mRNA levels were measured. Quantitative PCR data were normalized to 36B4 mRNA for all experiments. Data (mean ± se, n = 4) are expressed as a percentage of controls administered saline. *, P < 0.05; **, P < 0.01.

Fig. 6.

Effect of LPS in FGF21 KO mice. Wild-type (WT) and FGF21 KO mice were injected with LPS (5 mg/kg) and the animals were euthanized at 16 h. Plasma ketone (A) and free fatty acid (FFA) expressed as mean ± se (n = 7). *, P < 0.05; **, P < 0.01.

Fig. 7.

Effect of FGF21 on mouse survival. A, Survival curves of female C57Bl/6 mice subjected to LPS challenge and then treated with 30 μg of either recombinant human FGF21 protein or human albumin protein three times daily for 7 d starting at 1 h after LPS injection. B, Survival curves of female ob/ob mice receiving 50 μg of either recombinant human FGF21 or human albumin protein two to three times daily starting at 1 h after LPS injection. The data show survival up to 54 h after LPS challenge and are a composite of four independent experiments, each with similar results. C, Survival curves of female ob/ob mice receiving 50 μg of either recombinant FGF21 or human albumin protein two to three times daily for 7 d beginning at 1 h after LPS challenge. The data are a composite of two independent experiments each with similar results. D, Survival curves of female ob/ob mice subjected to CLP surgery and then treated with 50 μg of either recombinant human FGF21 protein or human albumin protein two times daily for 4 d starting at 2 h after CLP.