FGF-21: From Cloning to Clinic

The FGFs (fibroblast growth factors) are a family of molecules whose founding member (FGF basic) was discovered more than 65 years ago.1 Originally thought to be a family of mitogenic and differentiative molecules, two recent additions to the 22 member family are now generating excitement in the clinical community. The first molecule to attract attention was FGF-23, a 35 kDa polypeptide that seems to play a key role in promoting phosphorus excretion, independent of parathyroid hormone. Of particular interest is the ability to potentially reverse osteomalacia (calcium depletion) that accompanies abnormal phosphorus excretion.2-4 The second molecule of interest, and the focus of this article, is FGF-21. It is a liver-derived polypeptide that appears to have considerable potential for the treatment of diabetes mellitus.5, 6 In a recent paper, FGF-21 was found to act as an adipocyte-specific inducer of glucose uptake and to lower plasma triglyceride (TG) levels over an extended period.6 Notably, the effect is not immediate, and it is independent of insulin. FGF-21 effects on glucose uptake are additive, not synergistic with insulin. Moreover, unlike insulin, adipocyte responses to FGF-21 required exposure over a number of hours. The actual mechanism involved is unclear, but could involve a number of points along the glucose metabolic pathway (Figure 1).

Figure 1
Figure 1. Potential targets for FGF-21-mediated glucose uptake: FGF-21 may stimulate glucose uptake into adipocytes via FGF R modulation of adipocyte GLUT1. In addition, FGF-21 may enhance glucose uptake into glucagon-secreting pancreatic a-cells. In type II diabetics, this could have the effect of increasing insulin sensitivity by suppressing glucagon release, decreasing circulating glucose, and lowering the amount of insulin production required by the pancreatic ß-cells.

Normally, dietary glucose is absorbed into the intestinal vasculature and quickly encounters ß-cells of the pancreatic islets. Rodent ß-cells express GLUT2, a member of the SLC2 family of glucose and polyol transporters.7, 8 GLUT2 is unusual in that it is constitutively expressed on the cell surface and allows almost free diffusion of its target, glucose. Thus, any increase in extracellular glucose will be reflected by an almost immediate proportional increase in intracellular glucose. All rises in intracellular glucose are quickly followed by insulin release. The release is biphasic, peaking after three minutes, declining somewhat, and rising again after ten minutes for the duration of the glycemic episode.9 Released insulin encounters insulin receptors expressed on the principal targets of insulin such as muscle and fat. The first wave of insulin activates plasma membrane GLUT4 receptors, opening channels for glucose influx. The second and continuing wave of insulin induces GLUT4 translocation from internal vesicles to the plasma membrane, increasing the influx of glucose.10 Insulin resistance is a hallmark of type II diabetes, and is characterized by an inability to efficiently transport glucose into muscle and (white) fat. Approximately 75-90% of dietary glucose goes into muscle fibers, while 10% of plasma glucose is taken up by adipocytes.9, 11 GLUT4 is reportedly poorly expressed on muscle and fat in diabetes.8,10 This reduction could lead to hyperglycemia, since the “funnel” for glucose deposition would be reduced. GLUT4 would seem to be a possible target for FGF-21, an agent that causes glucose uptake.

Although it is tempting to speculate that FGF-21 might exert its glucose uptake effects via GLUT4, this doesn’t appear to be the case. Remarkably enough, FGF-21 seems to impact another GLUT transporter, GLUT1. GLUT1 activity seems to be independent of insulin action (at least on monocytes), and it is reported to be the predominant GLUT on human ß-cells (in contrast to rodent).7,12-14 FGF-21 is hypothesized to impact GLUT1 on adipocytes, but not skeletal muscle.6 The effect is probably indirect, as some isoform of FGF R1 and/or FGF R2 is likely to be the receptor for FGF-21.6 Although GLUT1 is a glucose transporter, it is unclear what effect FGF-21 could have on facilitated adipocyte glucose transport. Glucose entry into adipocytes generally results in its storage as TG. In the liver, plasma-derived glucose can be broken down to acetyl-CoA, and then reassembled from acetyl-CoA, two carbons at a time, into 16- and 18-carbon fatty acids. These can then be transported to the adipocyte via very low density lipoprotein (VLDL) where they are bound to glucose-derived, 3-carbon glycerol to form TG. In theory, this should result in increased TG stores and, by inference, enlarged adipocytes. However, FGF-21 transgenic mice, in which the human protein is over-expressed in the liver, exhibit white adipocytes that are smaller than normal. If FGF-21 does facilitate glucose influx, perhaps it does so on an expanded white adipocyte mass. Alternatively, adipocyte glucose may be metabolized and not used for fat storage.

FGF-21 has also been proposed to impact glucagon metabolism. In the fasting state, glucose levels are variable, maintained at a basal level by the opposing effects of insulin and glucagon. Glucagon is a hormone released by pancreatic islet a-cells in response to low glucose. It acts on its receptor, expressed by hepatocytes, to induce glucose release. Normally, after a meal, glucose levels are high, prompting insulin release and glucagon shutdown. In type II diabetes, however, glucagon would appear to be inappropriately expressed after a meal, promoting higher glucose levels than would otherwise be warranted.15 GLUT1 appears to be the glucose transporter in a-cells.16,17 In theory, a defective GLUT1 transporter in an environment of normo- or hyperglycemia could incorrectly signal hypoglycemia, with subsequent glucagon release. This would create abnormally high circulating glucose levels, and put pressure on the insulin-producing cells to release more insulin to correct the hyperglycemia (figure 1). A reduction in glucagon and plasma glucose could potentially lead to improved insulin sensitivity. If FGF-21 acts on GLUT1, it may be at the level of the a-cell. The insulin tyrosine kinase receptor is known to directly downregulate GLUT2 activity on hepatocytes, and a somewhat analogous situation may occur with tyrosine kinase FGF receptor(s).18


  1. Mohammadi, M. et al. (2005) Cytokine Growth Factor Rev. 16:107.
  2. Yamazaki, Y. et al. (2002) J. Clin. Endocrinol. Metab. 87:4957.
  3. Ferrari, S.L. et al. (2005) J. Clin. Endocrinol. Metab. 90:1519.
  4. Yamashita, T. et al. (2000) Biochem. Biophys. Res. Commun. 277:494.
  5. Nishimura, T. et al. (2000) Biochem. Biophys. Acta 1492:203.
  6. Kharitonenkov, A. et al. (2005) J. Clin. Invest. 115:1627.
  7. Zhao, L. et al. (2005) J. Endocrinol. 185:291.
  8. Uldry, M. & B. Thorens (2004) Pflugers Arch. Eur. J. Physiol. 447:480.
  9. Wilcox, G. (2005) Clin. Biochem. Rev. 26:19.
  10. Funaki, M. et al. (2004) Mol. Cell. Biol. 24:7567.
  11. Barnard, R.J. & J.F. Youngren (1992) FASEB J. 6:3238.
  12. Dimitriadis, G. et al. (2005) Cytometry Pt. A 64A:27.
  13. Thorens, B. (1996) Am. J. Physiol. 270:G541.
  14. De Vos, A. et al. (1995) J. Clin. Invest. 96:2489.
  15. Jiang, G. & B.B. Zhang (2003) Am. J. Physiol. Endocrinol. Metab. 284:E671.
  16. Heimberg, H. et al. (1995) J. Biol. Chem. 270:8971.
  17. Heimberg, H. et al. (1996) Proc. Natl. Acad. Sci. USA 93:7036.
  18. Eisenberg, M.L. et al. (2005) Cell. Physiol. Biochem. 15:51.