Correspondence: Karen Lam, Department of Medicine, University of Hong Kong, Queen Mary Hospital, 102 Pokfulam Road, Pokfulam, Hong Kong, China. Tel.: +852 2255-4783 / +852 2255-5323; Fax: +852 2816-2863 / +852 2904-9443; E-mail: email@example.com
Fibroblast growth factor 21 (FGF21), a metabolic hormone predominantly produced by the liver, is also expressed in adipocytes and the pancreas. It regulates glucose and lipid metabolism through pleiotropic actions in these tissues and the brain. In mice, fasting leads to increased PPAR-α mediated expression of FGF21 in the liver where it stimulates gluconeogenesis, fatty acid oxidation, and ketogenesis, as an adaptive response to fasting and starvation. In the fed state, FGF21 acts as an autocrine factor in adipocytes, regulating the activity of PPAR-γ through a feed-forward loop mechanism. Administration of recombinant FGF21 has been shown to confer multiple metabolic benefits on insulin sensitivity, blood glucose, lipid profile and body weight in obese mice and diabetic monkeys, without mitogenic or other side effects. Such findings highlight the potential role of FGF21 as a therapeutic agent for obesity-related medical conditions. However, in human studies, high circulating FGF21 levels are found in obesity and its related cardiometabolic disorders including the metabolic syndrome, type 2 diabetes, non-alcoholic fatty liver disease and coronary artery disease. These findings may indicate the presence of FGF21 resistance or compensatory responses to the underlying metabolic stress, and imply the need for supraphysiological doses of FGF21 to achieve therapeutic efficacy. On the other hand, serum FGF21 has been implicated as a potential biomarker for the early detection of these cardiometabolic disorders. This review summarizes recent developments in the understanding of FGF21, from physiological and clinical perspectives.
Since the discovery of the first fibroblast growth factor (FGF) almost 40 years ago, the FGF family has expanded over the years and currently consists of 22 members with a wide range of biological functions including cell growth, angiogenesis, wound healing and metabolism.[1, 2] The role of the FGFs in metabolism has been increasingly recognized in recent years, especially with the cloning and functional characterization of the FGF19 subfamily (FGF19, FGF21 and FGF21). Unlike the classical FGFs which require heparin for efficient binding to the FGF receptors (FGFRs) and act in a paracrine or autocrine fashion, FGF21 and the other endocrine FGFs lack the conventional heparin-binding domain and are secreted into the circulation, being able to escape the binding to the rich tissue depots of heparansulphate proteoglycans. Fibroblast growth factor 21 (FGF21) was first cloned and identified from mouse embryos by homology-based PCR in 2000. Human FGF21 is a polypeptide of 181 amino acids with 75% identity to mouse FGF21. It is secreted predominantly by the liver but also by other tissues involved in glucose and lipid metabolism such as the adipose tissue, pancreas and skeletal muscle. Studies in rodents have suggested FGF21 to be a key physiological regulator of fasting response, as well as a fed-state autocrine factor regulating the activity of PPAR-γ in adipose tissues. Administration of recombinant FGF21 in animal models (Fig. 1), including diabetic monkeys and findings in FGF21 transgenic mice have revealed potent in vivo beneficial effects of FGF21 on glucose and lipid metabolism, insulin sensitivity and body weight. Furthermore, unlike many of the other FGFs, FGF21 does not have effects on cell proliferation and tumourigenesis.[7, 8] Instead, over-expression of hepatic FGF21 delays the initiation of chemically induced hepatocarcinogenesis.
The favourable effects observed in animal studies would support the potential role of FGF21 as a therapeutic agent for diabetes and obesity. However, high serum levels of FGF21 are found in obese subjects and patients with disorders related to obesity and insulin resistance.[10, 11] The causes and underlying pathophysiology of elevated serum FGF21 in these pathological conditions warrant further clarification, although FGF21 resistance has been demonstrated in a study of obese mice. Whereas these observations imply that supraphysiological doses of FGF21 might be required for the treatment of such disorders in humans, they also suggest the potential use of FGF21 as a biomarker of obesity-related disorders.
Physiological roles of FGF21
FGF21: Production and regulation
The major site for FGF21 production is the liver. In mice, hepatic expression and circulating levels of FGF21 are raised by both fasting (for 12 h) and ketogenic diet and rapidly suppressed by refeeding. The nuclear receptor PPAR-α plays an indispensable role in fasting-induced hepatic expression of FGF21.[5, 13, 14] The mRNA expression of FGF21 in mouse livers and human primary hepatocytes are strongly induced by fenofibrate, a PPAR-α agonist. On the other hand, both fasting and fenofibrate-induced FGF21 expression are abolished in the absence of PPAR-α action, as demonstrated by experiments in PPAR-α KO mice. Apart from the liver, the adipocytes also express and secrete FGF21. At times of thermogenic activation, brown adipose tissue, in addition to being a FGF21 target, also becomes a source of systemic FGF21. This response is mediated by a powerful cAMP-mediated pathway, which regulates FGF21 gene transcription in response to noradrenergic stimulation. Studies in our laboratory have demonstrated differentiation-dependent expression of FGF21 and its enhancement by PPAR-γ activation in both 3T3-L1 murine adipocytes and human adipocytes. In addition, the degree of FGF21 expression in several types of adipose tissue is markedly raised in obese mice and becomes comparable to its expression in the liver. On the other hand, it has recently been shown to be an autocrine factor in the fed state, regulating the activity of PPAR-γ in adipose tissues through a feed-forward loop mechanism. FGF21 KO mice have defects in PPAR-γ signalling including decreased body fat and attenuation of PPAR-γ-dependent gene expression, and are refractory to the effects of the PPAR-γ agonist rosiglitazone, including both the beneficial insulin-sensitizing effects and the detrimental effects of weight gain and oedema. The changes in FGF21-KO mice are accompanied by a marked increase in the sumoylation of PPAR-γ, suggesting that FGF21 enhances PPAR-γ activity via posttranslational modifications. In summary, FGF21 can be secreted as an endocrine factor to co-ordinate the adaptive response to starvation or fasting; or as an autocrine factor induced in adipose tissue during the fed state to regulate adipocyte function.
Relatively little data are available on the regulation of FGF21 in humans as compared with mice. A positive correlation between FGF21 mRNA subcutaneous fat expression and its circulating level is found in the clinical samples of our study. In a human study, ketosis induced by fasting for 2 days or feeding a ketogenic diet is not associated with increased serum FGF21 levels, and unlike reports in mice, a significant increase in serum FGF21 levels is seen only after fasting for 7 days. However, in the same study, treatment with fenofibrate, a PPAR-α ligand, in patients with hypertriglyceridaemia results in increased FGF21 levels, suggesting that PPAR-α also regulates FGF21 in humans. In line with such an observation, another group has also found that the in vitro expression of human FGF21 gene is increased by two fasting signals, namely PPAR-α and glucagon-PKA. On the other hand, serum FGF21 levels and FGF21 mRNA expression in visceral fat are increased in subjects with obesity, a condition associated with over-nutrition.[10, 18] It is possible that FGF21 is induced in extreme nutritional conditions, including prolonged fasting or over-feeding, in humans.
FGF21 and glucose metabolism
The metabolic activity of FGF21 was first discovered in a high throughput screen using 3T3-L1 adipocyte glucose uptake as an assay system. FGF21 was found to activate glucose uptake in adipocytes, an effect independent of insulin and observed after at least 4 h of treatment, in contrast to the rapid action of insulin. A possible explanation is that insulin acts through GLUT4 translocation whereas FGF21 acts via GLUT1 mRNA up-regulation.
Chronic systemic administration of FGF21 to mice with genetic obesity leads to amelioration of fasting hyperglycaemia via increased glucose disposal and improved hepatic insulin sensitivity. Interestingly, chronic intraventricular infusion of FGF21 for 2 weeks in rats also results in increased insulin sensitivity due to an enhanced insulin-induced suppression of hepatic glucose production and gluconeogenic gene expression, with no change in glucose utilization, suggesting that the beneficial effect of FGF21 on hepatic insulin resistance may be, at least in part, mediated via central pathways. Body weight in these rats remains unchanged as both food intake and energy expenditure are increased. These findings may have therapeutic relevance as FGF21 crosses the blood brain barrier in a nonsaturable, unidirectional manner.
FGF21 is expressed in the pancreas, as is the single-pass transmembrane protein β-Klotho, an important component of the FGF21 receptor complex, which determines the tissue selectivity of FGF21 action. In isolated rat pancreatic islets, FGF21 inhibits glucagon secretion and increases insulin mRNA and proteins, but enhances glucose-induced insulin secretion only in islets from diabetic rodents. Rat islets and INS-IE cells treated with FGF21 are partially protected against glucolipotoxicity, probably through improved β-cell function and survival, via the activation of ERK 1/2 (extracellular signal-regulated kinase 1/2) and Akt signalling pathways.
FGF21 may also contribute to the glucose-lowering action of PPAR-γ agonists. Treatment of 3T3-L1 adipocytes with FGF21 and rosiglitazone, a PPAR-γ agonist, in combination leads to a synergistic increase in glucose transport. This suggests a profound functional synergy between the FGF21 and PPAR-γ pathways. Whereas FGF21 can enhance the transcription activity of PPAR-γ, the expression of β-Klotho is stimulated by rosiglitazone.
On the other hand, FGF21 has also been implicated in the regulation of gluconeogenesis as fasting progresses to starvation but, unlike glucagon, it does not stimulate glycogenolyis. Its effect is mediated by the induction of hepatic expression of peroxisome proliferator-activated receptor co-activator protein 1 α (PGC1α), a transcriptional co-activator controlling the expression of gluconeogenic genes. Mice lacking FGF21 fail to fully induce PGC1α expression in response to fasting and have impaired gluconeogenesis. However, FGF21 may also act directly on the liver to stimulate the expression of gluconeogenic genes, as suggested by the finding that FGF21 can stimulate the same degree of gluconeogenic gene expression in a study of mice with liver-specific ablation of PGC1α.
Thus, based on the aforementioned animal studies, FGF21 can impact glucose metabolism via multiple mechanisms, acting through its receptor complex in the liver, adipose tissue, brain and pancreas (Fig. 2).
FGF21 and lipid metabolism
Different studies have demonstrated that FGF21 is required for ketogenesis in mice in the fasting state. Transgenic mice with liver-specific overexpression of FGF21 exhibit a significant increase in serum ketone bodies and a concurrent reduction in serum and hepatic triglyceride concentrations. However, inconsistent results have been reported regarding the physiological role of endogenous FGF21 in ketogenesis. FGF21 KO mice in one study demonstrate impaired adaptation to ketosis induced by a ketogenic diet, whereas a Japanese group has reported increased ketogenesis in FGF21 KO mice fasted for 24 h, as evidenced by a modest increase in serum β-hydroxybutyrate levels.
On the other hand, a human study shows no correlation between plasma levels of FGF21 and ketone bodies after a 2-day fast or feeding with a ketogenic diet. In another human study, neither fasting up to 72 h nor a ketogenic diet for 12 days increases serum FGF21 levels. With these conflicting findings, the physiological role of FGF21 in regulating ketogenesis remains unclear.
Despite early data suggesting that acute treatment with recombinant FGF21 increases lipolysis, more recent studies have shown an inhibitory effect. The observation of an increase in non-esterified fatty acids in adenovirus-mediated FGF21 knockdown mice on a ketogenic diet is in line with the ability of FGF21 to inhibit lipolysis. More recently, we have also shown that FGF21 can suppress growth hormone (GH)-induced lipolysis in mice through a feedback regulatory loop. GH is released from the pituitary in response to fasting and stimulates lipolysis in fat cells. The resulting increase in circulating free fatty acids (FFAs) induces hepatic FGF21 production via the action of PPAR-α. Raised FGF21 in turn feedbacks negatively to terminate GH-induced lipolysis in adipocytes. The greater rise in serum glycerol and FFAs in response to GH in the FGF21 KO mice, compared with their wild-type littermates, also supports the inhibitory effect of endogenous FGF21 on lipolysis.
In human adipocytes, FGF21 attenuates lipolysis stimulated by catecholamine and atrial natriuretic peptide after treatment for three days. Human data from our group have also shown that the 24-h profiles of FFAs correlate closely with those of FGF21. A strong positive association is found between the peak levels of circulating FFAs and FGF21 during both day and night, with the peak time of FFAs preceding that of FGF21 by 3–4 h. These findings also suggest the existence of feedback regulation between FFAs and FGF21 and a role of endogenous FGF21 in suppressing excessive lipolysis in humans.
FGF21 plays a part in the induction of hepatic fatty acid oxidation by PPAR-α. In mice on a ketogenic diet, studies based on hepatic FGF21 knockdown show that FGF21 is required for the normal activation of hepatic lipid oxidation and triglyceride clearance. Chronic treatment with recombinant FGF21 reduces serum and hepatic triglyceride levels, and reverses fatty liver disease in diet-induced obese mice, through the inhibition of SREBP-1 (sterol regulatory element binding protein-1), a transcription factor critical for lipogenesis. In diabetic monkeys treated with FGF21, reductions in serum triglycerides, cholesterol and small dense LDL-cholesterol, together with increases in HDL-cholesterol (Fig. 1), are observed.
The reduction in circulating FFAs, consequent to the inhibition of excess lipolysis and enhanced hepatic fatty acid oxidation, may contribute the reduction in systemic insulin resistance in the FGF21-treated obese and diabetic animals. In this context, the reduction in hepatic steatosis by FGF21 can also lead to an amelioration of hepatic insulin resistance.
FGF21 and clinical conditions
FGF21 in obesity, type 2 diabetes and other disorders with insulin resistance
Despite the multiple beneficial effects of FGF21 on insulin sensitivity, glucose and lipid homeostasis in animal models,[7, 8] we made the surprising observation that raised circulating FGF21 levels were present in obese diabetic db/db mice as well as in obese/overweight humans. In the db/db mice, increased FGF21 gene expression was found in both the liver and adipose tissue. In the human subjects serum FGF21 correlated positively with adiposity, fasting insulin and triglycerides, but negatively with HDL-cholesterol, after adjusting for age and BMI. An independent association was found between serum FGF21 levels and the metabolic syndrome in adults but not in children. In those children, however, serum FGF21 was also raised in obesity and correlated with FFA and leptin. Successful weight reduction following one year of lifestyle intervention in these children was accompanied by a significant reduction in FGF21, accompanied by a reduction in serum FFA, insulin and insulin resistance index, suggesting that FGF21 is a result rather than the cause of obesity and may have occurred in compensation to the metabolic stress, such as raised circulating FFA and insulin levels, in states of obesity and insulin resistance. The stimulation of FGF21 by FFA, via PPAR-α activation, is well established.[5, 13, 14, 35] Lipid infusion in humans induces an elevation in serum FGF21 levels, with a strong correlation being found between the changes in FGF21 and FFA levels. Hyperinsulinaemia during a euglycaemic clamp also increases FGF21 levels in insulin-resistant humans with impaired glucose tolerance. In line with this hypothesis, raised circulating FGF21 levels in patients with impaired glucose tolerance and type 2 diabetes[11, 36] correlate inversely with whole-body insulin sensitivity and directly with hepatic insulin resistance, and the reduction in FGF21 following intensive insulin therapy of type 2 diabetic patients correlates positively with the amelioration of insulin resistance. We made the first observation that high plasma levels of FGF21 in 1292 non-diabetic subjects at baseline significantly predicted diabetes development over 5·4 years, even when traditional risk factors were taken into consideration. Whether FGF21 resistance predisposes to the development of diabetes remains to be investigated.
Elevated serum levels of FGF21 are also found in subjects with other insulin resistant states (Table 1), including dyslipidaemia and coronary artery disease, non-alcoholic fatty liver disease (NAFLD)[29, 39] and polycystic ovarian syndrome. FGF21 resistance has been proposed as one of the causes for the raised FGF21 circulating levels in obese mice. Treatment of mice with diet-induced obesity with FGF21 resulted in both a significantly attenuated signalling response as assessed by (ERK1/2) phosphorylation as well as an impaired induction of FGF21 target genes, including cFos and EGR1. A recent study suggests that adipose tissue inflammation in obesity, involving the JNK1 pathway, can lead to the repression of β-Klotho expression by TNF-α and hence impaired FGF21 action in adipocytes. Similar mechanisms may also lead to FGF21 resistance in other conditions with chronic subclinical inflammation such as the metabolic syndrome, type 2 diabetes, coronary artery disease and NAFLD.
Table 1. SerumFGF21 levels in different clinical conditions
TG, triglycerides; HDL-C, HDL-cholesterol; BMI, body mass index; SDS-BMI, standard deviation score of body mass index; FFA, free fatty acid; IGT, impaired glucose tolerance; T2DM, type 2 diabetes; LDL-C, LDL-cholesterol; SBP, systolic blood pressure; IR, insulin resistance; NAFLD, non-alcoholic fatty liver disease; PCOS, polycystic ovary syndrome; HOMA-IR, homeostasis model assessment of insulin resistance; GH-AUC, Area under the curve of growth hormone; T1DM, type 1 diabetes; LADA, latent autoimmune diabetes in adults; LVMI, left ventricular mass index; UAE, urinary albumin excretion.
Positive correlation with adiposity, fasting insulin, TG; Negative correlation with HDL-C
The increased FGF21 gene expression in the liver of patients with NAFLD[29, 39] is in keeping with a compensatory response to increased lipid synthesis consequent to FGF21 resistance as FGF21 is known to inhibit SREBP-1, a transcription factor critical for lipogenesis, which has been implicated in hepatic steatosis. FGF21 resistance is also evident in obese human subjects who display an impaired nocturnal rise and reduced circadian rhythmicity in circulating FGF21, probably resulting from the desensitizing effect of the high daytime FGF21 levels.
FGF21 and anorexia nervosa
Because of the known nutritional regulation of FGF21 in mice,[5, 13] changes in serum FGF21 levels have been investigated in patients with anorexia nervosa (AN), a condition of chronic severe malnutrition. Conflicting results have been obtained.[42, 43] Dostalova et al. found that patients with a severe restrictive subtype of AN had significantly lower plasma FGF21 levels than the control group, suggesting that plasma FGF21 levels in humans are not increased with chronic malnutrition, as distinct from the rise in serum FGF21 levels in response to fasting in animal studies.[5, 13] However, these findings in adult women with severe AN may be confounded by a marked degree of reduced adiposity as FGF21 levels were closely related to BMI, leptin and adiponectin in both AN subjects and controls. Fazeli et al., on the other hand, showed that FGF21 levels were elevated in adolescent girls with AN compared with healthy controls after controlling for body fat percentage and insulin resistance. Circulating Levels of FGF21 of the two groups were similar before the adjustments were made. AN patients have significantly lower fat depots compared with controls, which may result in a decrease in FGF21 production from adipocytes, a significant source of serum FGF21 levels in humans. However, elevated liver-derived FGF21 in response to chronic malnutrition may offset the decrease in fat-derived FGF21, and the difference observed in the two studies might be explained by the greater reduction in body fat in the AN patients studied by Dostalova et al. Serum IGF-1 levels were reduced in AN patients in the study by Fazeli et al. and a strong inverse association was found between serum FGF21 and IGF-1 levels. These findings, together with a positive correlation between serum FGF21 and the integrated nocturnal GH levels, suggest that FGF21 elevation may contribute to GH resistance in AN. FGF21 has been shown to induce GH resistance through reduced phosphorylation of STAT5 (signal transducer and activator of transcription-5) in FGF21 transgenic mice.
FGF21 and autoimmune diabetes
Serum levels of FGF21 show distinct changes in different subtypes of diabetes as shown by our previous report comparing serum FGF21 levels in patients with type 1 diabetes, latent autoimmune diabetes in adults (LADA) and type 2 diabetes. Serum FGF21 level is increased in type 2 diabetes, but decreased in type 1 diabetes and LADA, compared with age- and weight-matched controls. In type 1 and LADA patients, serum FGF21 levels correlate positively with serum C-peptide levels, but inversely with titres of autoantibodies against glutamic acid decarboxylase and insulinoma-associated protein 2. The decreased FGF21 levels in these two subtypes of autoimmune diabetes might reflect on the impairment in pancreatic beta cell function. FGF21 and its receptor complex are expressed in pancreatic islets as well as in rat primary β-cells and INS-1E cells.[4, 23] FGF21 KO mice are more prone to cerulean-induced pancreatitis, whereas FGF21 transgenic mice are relatively resistant to this condition. The decreased circulating levels of FGF21 in autoimmune diabetes might be attributed to reduced FGF21 production in the pancreas, consequent to the immunologically mediated tissue injury, which in turn further aggravates islet β-cell destruction as a result of the reduction in this protective autocrine factor.
FGF21 and kidney diseases
Serum FGF21 levels have been shown to be increased in patients with impaired renal function. Patients undergoing chronic haemodialysis have elevated serum FGF21 levels, more than 15-fold that of controls, while serum creatinine and GFR are inversely related to circulating FGF21 levels in control subjects. FGF21 is eliminated by the kidneys and its level increases as the stage of chronic kidney disease progresses. FGF21 is also independently related to urinary albumin excretion in type 2 diabetes, not only in those with macroalbuminuria, but also in patients with microalbuminuria, suggesting that FGF21 may be regarded as an early indicator of subclinical diabetic nephropathy. While the high circulating FGF21 levels in subjects with overt diabetic nephropathy is likely a result of decreased renal clearance, its increase in early diabetic nephropathy may be partly attributed to a compensatory response to the underlying metabolic disturbance such as dyslipidaemia and enhanced insulin resistance.
Other clinical studies on FGF21 circulating levels
In the past 4 years, there have been extensive studies on the circulating levels of FGF21 in various clinical conditions (Table 1), which have improved our understanding of the possible regulation and actions of FGF21 in humans. It should be noted that circulating levels of FGF21 represent the integrated result of secretion from the different tissues of production and its clearance, and are also influenced by secondary changes in response to other confounding conditions such as metabolic stress and FGF21 resistance. In addition to potential species differences in FGF21 regulation and actions, findings from animals and cell-based studies (even those performed on human cells) may not be reproduced in clinical studies. For example, PPAR-α ligands such as treatment with fenofibrate stimulate FGF21 expression in human hepatocytes and raise serum FGF21 levels in humans,[16, 18] but the in vitro effect of rosiglitazone, a PPAR-γ agonist which stimulates FGF21 expression in human adipocytes, cannot be reproduced in vivo. Treatment with rosiglitazone in type 2 diabetic patients results in a reduction in serum FGF21 levels, probably secondary to a reduction in insulin resistance.
On standard meals, the 24-h oscillatory pattern of circulating FGF21 levels are opposite to those of serum glucose and insulin, but resemble those of FFA and cortisol. Our correlation analysis, together with in vitro studies in human hepatocytes, suggests that the circadian rhythm of circulating FGF21 is caused in part by oscillation in FFA. These findings provide strong support for the role of FGF21 as an important metabolic regulator that integrates the circadian rhythm with energy homeostasis in humans. Whether glucocorticoids are involved in the regulation of FGF21 expression is not known. Although raised serum FGF21 levels have been reported in patients with Cushing's Syndrome, this may just reflect on the increased adiposity and metabolic disturbance as no significant correlation between serum FGF21 and cortisol levels can be demonstrated. The inverse relationship of serum FGF21 with acute changes in serum glucose and insulin under physiological conditions is also seen during oral glucose tolerance tests. On the other hand, chronic hyperglycaemia[11, 18] and hyperinsulinaemia are associated with high serum FGF21 levels.
FGF21 and its pharmacological effects
Therapeutic administration of recombinant FGF21 has led to various beneficial effects in animal models, including lowering of blood glucose and triglyceride levels, reversing hepatic steatosis and improving insulin resistance.[7, 33] The effect of FGF21 lasts at least 24 h and with no weight gain, mitogenicity or hypoglycaemia being observed in the FGF21 treated animals.[7, 8] Chronic infusion of FGF21 for 8 weeks in db/db mice almost normalizes their glucose levels, in part contributed by the effects on β-cell survival and function in the diabetic animals. This is of potential therapeutic importance as progressive β-cell failure is a major clinical challenge in the treatment of type 2 diabetes.
Despite the presence of FGF21 resistance in obese mice and possibly in humans, beneficial metabolic effects have been clearly observed in obese mice[7, 54] and diabetic monkeys after the administration of FGF21. Systemic administration of FGF21 for 2 weeks in mice with diet-induced or genetic obesity lowers their mean body weight by 20% predominantly via a reduction in adiposity, with no change in total caloric intake or physical activity. Improved hepatic steatosis is also observed. FGF21-treated obese mice also show increased energy expenditure and a lower respiratory quotient, reflecting preferential utilization of fat as a fuel source.
The above observations in animal studies would suggest FGF21 as a potential candidate for consideration as a novel treatment for metabolic conditions related to obesity, including diabetes and dyslipidaemia. The findings in the clinical studies also support the role of FGF21 as an emerging metabolic hormone in humans, especially with regard to lipid metabolism. The raised FGF21 levels observed in patients with obesity and related cardiometabolic disorders may appear paradoxical, but are nevertheless reminiscent of the presence of hyperinsulinaemia in patients with type 2 diabetes in whom the use of insulin remains an effective treatment, despite the presence of insulin resistance. The therapeutic responses to FGF21 in these conditions, particularly in patients already having high circulating FGF21 level, remain to be investigated. The short circulating half-life of FGF21 can be potentially overcome by the use of PEGylation. However, it is noteworthy that the therapeutic doses of recombinant FGF21 used in these animal studies are supraphysiological. The necessity to use such a high dose may be due to the low bioavailability of the recombinant FGF21 or the FGF21 resistance in obese/diabetic animals. An alternative therapeutic strategy would be the development of FGF21 sensitizing agents, analogous to the use of insulin sensitizers for treating type 2 diabetes.
A recent study reported the role of FGF21 in the regulation of skeletal homeostasis. FGF21 was shown to inhibit osteoblastogenesis and stimulate adipogenesis from bone marrow mesenchymal stem cells by potentiating the activity of PPAR-γ. Both genetic and pharmacological FGF21 gain of function suggested a role of FGF21 in inducing a remarkable decrease in bone mass. At this juncture, whether skeletal fragility and increased fracture risk may be an undesirable consequence of chronic FGF21 administration remains to be confirmed by other investigators. Nonetheless, while the therapeutic efficacy and safety of FGF21 in the treatment of obesity and its related disorders are being actively investigated, further research may reveal the possibility of developing tissue or pathway selective FGF21 agonists as another strategy for FGF21-based therapy.
FGF21 mediates the crosstalk between different metabolic organs to regulate glucose and lipid metabolism. It is regulated by PPAR-α in liver and is linked to the biology of fasting and ketogenesis. FGF21 also plays a part in the fed state, functioning as an autocrine factor in adipose tissue, regulating the activity of PPAR-γ. Therapeutic administration of recombinant FGF21 has led to favourable metabolic effects on body weight and various cardiometabolic risk factors in animal models. In human studies, individuals with obesity and its related cardiometabolic disorders including metabolic syndrome, type 2 diabetes, NALFD and coronary artery disease are associated with high serum FGF21 levels, which may be due to a compensatory response to protect the body from adverse metabolic conditions. The findings in human studies suggest that serum FGF21 level has the potential to be an important biomarker for the early diagnosis of metabolic diseases or its complications. The potential of FGF21 as a novel therapeutic target for obesity and its related clinical disorders warrants further exploration.
Studies on FGF21 at the University of Hong Kong were supported by CRF 03/09M from the Hong Kong Research Grant Council.