Glucocorticoids and Diabetes

Abstract

Glucocorticoids (GCs), stress hormones produced by the adrenal gland, are involved in many pathways in physiology and metabolism including glucose homeostasis and inflammation. Excess GC signaling results in Cushing’s syndrome and possibly metabolic syndrome. Diabetes, central adiposity, and hyperlipidemia are components of both syndromes. Here, we discuss the mechanisms of GC action, clinical syndromes of GC excess, modulation of glucose homeostasis by GCs, and future treatments for diabetes based on GC signaling.

Introduction

Type 2 diabetes mellitus (DM) affects nearly ten percent of the population of the United States and is associated with a host of microvascular and macrovascular complications. Healthcare expenditures aimed at the management of diabetes and its complications exceed $245 billion yearly^1–3. As the incidence of diabetes continues to rise, there has been a major effort to identify novel molecular targets that may be of therapeutic benefit.

Obesity is a strong risk factor for the development of insulin resistance, a process that is central to the underlying pathophysiology of DM. As insulin resistance increases, the pancreatic β-cell is initially able to compensate and augment insulin secretion to maintain euglycemia. Eventually, pancreatic β-cell failure leads to a deficit of insulin and the onset of DM. The underlying mechanisms of insulin resistance continue to be an active area of research. Adipocyte dysfunction in response to chronic nutrient overload has been implicated. In lean individuals, free fatty acids (FFA) are sequestered in the form of triglyceride (TG) during periods of caloric excess and released to meet energy demands by the process of lipolysis. In contrast, the adipocyte of obese individuals is dysfunctional, in large part due to the inflammatory milieu that accumulates in adipose tissue in response to prolonged fuel-storage efforts. This was first observed in mouse models of diet-induced obesity⁴ and soon after confirmed in the human condition⁵. The end-result is impaired FFA storage, accumulation of ectopic lipid, and rising serum levels of FFAs and inflammatory cytokines, resulting in systemic insulin resistance⁶.

Glucocorticoids (GCs) are stress hormones involved in the regulation of glucose homeostasis, adipocyte development, and inflammation. Clinical syndromes of glucocorticoid excess are characterized by the development of diabetes and visceral adiposity in a majority of cases, and mouse models of localized adipocyte-specific GC excess develop visceral adiposity and insulin resistance^7,8. In this review, we will further explore the mechanisms regulating glucocorticoid production and metabolism, the clinical and basic science literature supporting a role for glucocorticoids in the pathogenesis of DM, and their potential role as a therapeutic target in DM.

Glucocorticoid Regulation and Action

Glucocorticoids are steroid hormones produced by the adrenal cortex. Circulating levels of glucocorticoids are regulated by the hypothalamic-pituitary-adrenal (HPA) axis, a neuroendocrine feedback circuit, whereby production of corticotropin releasing hormone (CRH) by the hypothalamus drives production of adrenocorticotropic hormone (ACTH), which in turn stimulates the adrenal gland to synthesize cortisol, the primary active GC in humans. The HPA axis is activated in response to stress, circadian rhythms, and other acute stimuli. Circulating GCs feedback at the level of the hypothalamus and pituitary to suppress the production of CRH and ACTH and subsequent synthesis and release of GCs from the adrenals. Only about 5% of circulating cortisol is in the free, bioactive form. The remainder is bound to cortisol-binding globulin (CBG) and albumin⁹.

The effects of glucocorticoids are mediated by the glucocorticoid and mineralocorticoid receptors (GR and MR). GCs and mineralocorticoids bind MR with equal affinity, but GCs circulate at much higher concentrations than mineralocorticoids (aldosterone). How then do mineralocorticoid responsive tissues retain sensitivity to the less abundant aldosterone? Tissue-specific regulation of GCs is achieved by 11β-hydroxysteroid dehydrogenases (11βHSD). 11βHSD2 is expressed primarily in mineralocorticoid responsive tissues such as the kidney and catalyzes the inactivation of cortisol to cortisone, preventing excessive activation of MR by GCs and facilitating activation of MR by the less abundant ligand, aldosterone. 11βHSD1, on the other hand, is expressed primarily in metabolically active tissues implicated in the pathophysiology of metabolic syndrome such as liver and adipose, and catalyzes the converse reaction. GCs exert the majority of their effects on glucose metabolism through activation of GR^9,10.

GR is a member of the nuclear hormone receptor family of transcription factors. Binding of GCs to GR results in dissociation of GR from its cytosolic HSP90 complex and promotes translocation to the nucleus where GR regulates the transcription of GR target genes. Activation of transcription is induced by GR binding to GR response elements (GREs) and interaction with transcriptional coactivators. Repression of transcription is induced by GR binding to GREs with subsequent interaction with transcriptional corepressors or by the DNA-independent direct interaction of GR with other transcription factors, a process known as tethering¹¹.

Clinical and Subclinical Glucocorticoid Excess

Chronic glucocorticoid excess is associated with the development of Cushing’s Syndrome (CS). Clinically, this disease is characterized by central adiposity, muscle atrophy, loss of bone mass, hyperpigmented abdominal striae, skin thinning, neuropsychological disturbances, resistant hypertension, and insulin resistance/diabetes. The biochemical diagnosis of CS is often challenging and relies on the confirmed loss of the diurnal variation in cortisol levels (as assessed by midnight salivary cortisol measurements), impaired attenuation of cortisol production in response to exogenous glucocorticoids (as assessed by an overnight dexamethasone suppression test), and/or measurement of frankly elevated 24-hour urinary free cortisol levels^12,13. Further discussion of the diagnosis of CS is beyond the scope of this review.

GCs have long been used for their anti-inflammatory properties in the treatment of a number of common autoimmune disorders, and long-term exogenous GC therapy continues to be the most common cause of Cushing’s syndrome. Cushing’s syndrome related to exposure to endogenous GCs may be either ACTH-dependent or -independent. In the vast majority of cases, ACTH-dependent CS is the result of an ACTH-secreting pituitary adenoma. Rarely, cases are the result of an ACTH-secreting neuroendocrine tumor, classically small cell lung cancer. ACTH-independent CS is less common and often secondary to direct production of cortisol by the adrenal gland (adenoma or hyperplasia)^12,13.

Diabetes is a common complication of CS. About 50% of individuals have altered glucose metabolism, classified as impaired fasting glucose, impaired glucose tolerance, or DM, at the time of CS diagnosis. Of these patients, two-thirds have DM¹⁴. Using fasting hyperglycemia to diagnose DM in patients with CS is limited. In one series, nearly two-thirds of all patients with CS had normal fasting blood glucoses. This finding highlights the effect of GCs in promoting post-prandial glucose excursions and suggests that hemoglobin A1C and oral glucose tolerance testing may be more useful in the diagnosis of GC-induced DM. However, there was a positive correlation between fasting plasma glucose and morning plasma cortisol levels, indicating that derangements in fasting glucose values are likely to be more common in individuals with more severe hypercortisolemia¹⁵. With exogenous GC therapy, the incidence of diabetes after initiation of GCs, as expected, is proportional to the dose and duration of GC administration. A case-control study demonstrated that patients were more than two times likely to develop new-onset DM after initiation of GC therapy compared to controls, as determined by the need for initiation of either insulin or other oral diabetic therapies. This effect is dose-dependent, as illustrated by odds ratios for the development of diabetes of 3.02, 5.82 and 10.35 for patients receiving the daily equivalent of 50, 100 and greater than 120 mg of hydrocortisone, respectively (a healthy person synthesizes the equivalent of 15 mg hydrocortisone per day)^16,17. The increased incidence of DM in CS undoubtedly contributes to the two-fold increase in mortality observed in affected patients¹².

As expected, in diabetic patients treated with exogenous glucocorticoids, insulin requirements increase significantly, and glycemic control often deteriorates. This observation holds true across the various etiologies of DM, including patients with type 1 DM, type 2 DM, and cystic fibrosis-related DM¹⁸. In hospitalized patients treated with at least 40 mg of prednisone, hyperglycemia, as defined by a blood glucose of greater than 200 mg/dl, occurred in 56% of patients, and nearly two-thirds of those patients experienced recurrent hyperglycemia¹⁹. In a study utilizing continuous glucose monitoring to assess circadian patterns of glycemic control, hyperglycemia occurred more frequently in the afternoon and early evening, providing some rational to the use of increased prandial insulin coverage or the addition of NPH insulin, which has a peak activity at 6–8 hours, to the patient’s insulin regimen^18,20,21. The degree of increase in insulin requirements is difficult to quantify and is dependent on steroid dose and potency. In one study of type 1 diabetic patients receiving 60 mg of prednisone (equivalent to ~240 mg hydrocortisone) daily for 3 days, insulin requirements increased by an average of 70%²².

CS is a rare condition with an estimated incidence of approximately two individuals per million in the United States. However, the description of metabolic syndrome (MetS)²³, a condition that affects nearly one-quarter of adults and is characterized by obesity, hyperlipidemia, and insulin resistance²⁴, prompted the realization that MetS may represent a subset of patients with CS or subclinical CS (SCS). SCS is most commonly associated with cortisol-producing adrenal adenomas rather than pituitary Cushing’s disease and is characterized by mild hypercortisolism and metabolic abnormalities, including obesity, hypertension, and insulin resistance, without the classical clinical signs of CS. Remarkably, the incidence of clinical hypercortisolism in individuals with diabetes or MetS has been reported to approach 3–10% of those screened²⁵.

Studies testing the hypothesis that MetS represents a milder form of overt Cushing’s syndrome have failed to demonstrate strong associations between circulating cortisol levels and features of MetS. While some have observed subtle changes in HPA axis activity in metabolic syndrome patients, these do not fully explain the etiology of MetS²⁶. However, the description of tissue-specific conversion of inactive GCs to active GCs by 11βHSD1 in tissues prone to GC-induced insulin resistance has rejuvenated interest in GCs as the underlying etiology of metabolic syndrome. Indeed, rodent models employing targeted overexpression of 11βHSD1 in the adipocyte demonstrated a phenotype similar to that of metabolic syndrome, characterized by insulin resistance, hyperlipidemia, and visceral adiposity⁸. Furthermore, human studies have shown a correlation between expression levels of adipocyte 11βHSD1 and intra-abdominal fat as well as impaired insulin sensitivity²⁷. These findings have led to the development of the so called “dehydrogenase hypothesis” to explain the tissue-specific amplification of GC action as a mechanism behind the development of tissue-specific Cushing’s syndrome and systemic metabolic syndrome^28,29.

Regulation of Glucose Metabolism by Glucocorticoids

Glucocorticoids impact all aspects of glucose metabolism by exerting their collective effects on the liver, endocrine pancreas, skeletal muscle, and adipose tissue to antagonize insulin action (See Figure 1). The consequences of GR activation highlight the catabolic effects of glucocorticoid excess favoring the liberation of energy stores. In the short-term, these effects are critical for adaption to acute stress and illness in an effort to preserve circulating glucose levels and facilitate continued brain function. However, as described above, the long-term consequences of GC excess include hyperglycemia and insulin resistance. The effects of GC excess in each target tissue will be discussed below as it pertains to glucose utilization, gluconeogenesis, glycogen metabolism, and pancreatic endocrine function³⁰.

Figure 1.

Functions of Glucocorticoids in Peripheral Tissues to Regulate Glucose Homeostasis.

Figure provided by Jen-Chywan Wang and adapted.

Gluconeogenesis occurs primarily in the liver and is the process by which glucose is generated from the non-carbohydrate substrates lactate, glycerol, and amino acids. An in depth discussion of this pathway is beyond the scope of this review, but the key GC-regulated steps will be described (See Figure 2). GCs are critical for the generation of gluconeogenic precursors by promoting lipolysis of triglyceride stores in adipose tissue and protein degradation in muscle, resulting in the production of glycerol and gluconeogenic amino acids, respectively. GCs also positively regulate gluconeogenesis by inducing binding of GR to the glucocorticoid response elements in the promoter region of several genes encoding enzymes involved in gluconeogenesis, including glucose-6-phosphatase and phosphoenolpyruvate carboxykinase, the rate limiting step of gluconeogenesis^30,31.

Figure 2.

Gluconeogenic Pathway in Hepatocytes. Gluconeogenic precursors (lactate and alanine) are converted to pyruvate, which enters the mitochondria and is converted to oxaloacetate (OAA) by pyruvate carboxylase. Through the malate-aspartate shuttle, OAA exits the mitochondria to form phosphoenolpyruvate (PEP). OAA may also be converted to PEP directly within the mitochondria. PEP then feeds into the gluconeogenic pathway. In addition, glycerol is metabolized to dihydroxyacetone phosphate (DHAP), which is then converted to fructose-1,6-bisphosphate (F1,6BP). The final product, glucose, is produced in the ER by the glucose-6-phosphatase catalytic subunit (G6PC). The key enzymes are boxed, with GR primary targets shown in yellow. Abbreviations: OAA oxaloacetate, PEP phosphoenolpyruvate, DHAP dihydroxyacetone phosphate, G3P glyceraldehyde-3-phosphate, F1,6BP fructose-1,6-bisphosphate, 2-PG 2-phosphoglycerate, 3-PG 3-phosphoglycerate, 1,3-BPG 1,3-bisphosphoglycerate, G3P glyceraldehyde-3-phosphate, F1,6BP fructose-1,6-bisphosphate, F6P fructose-6-phosphate, and G6P glucose- 6-phosphate. Enzyme abbreviations: PC pyruvate carboxylase, m-PCK1 mitochondrial phosphoenolpyruvate carboxykinase, PCK1 cytosolic phosphoenolpyruvate carboxykinase, FBP1 fructose-1,6- bisphosphatase 1, PFK1 phosphofructokinase 1, G6PC glucose-6-phosphatase catalytic subunit.

Figure provided by Jen-Chywan Wang and adapted.

Regulation of glycogen metabolism by GCs is distinct from gluconeogenesis in that it is tissue-specific. In the liver, GCs induce glycogen formation by increasing the activity of glycogen synthase^30,32. In contrast, GCs primarily impair insulin-stimulated glycogen synthesis in skeletal muscle by attenuating the activity of glycogen synthase^30,33.

Glucose utilization and oxidation takes place in white adipose tissue and skeletal muscle in response to insulin signaling. The mechanism by which GCs antagonize insulin signaling in the skeletal muscle is, at least in part, dependent on attenuation of insulin-induced GLUT4 translocation to the cell membrane^30,34. GCs also appear to promote insulin resistance in skeletal muscle by regulation of a number of GR target genes involved in the insulin-signaling cascade, resulting in an apparent post-receptor defect with reduced downstream phosphoninositide-3-kinase and AKT activities^35,36. GC-induced generation of lipid mediators in the form of fatty acids from the adipocyte and ceramides from the liver may also play a role in the development of systemic insulin resistance as these products accumulate in the liver and skeletal muscle^6,30,37. The mechanisms contributing to impaired glucose utilization by GCs in the adipocyte are less clear at this time but likely to involve modulation of GLUT4 function and the insulin signaling cascade³⁰. GCs are also involved in the differentiation and expansion of adipocyte precursors, a process that may further exacerbate insulin resistance and adipocyte dysfunction^30,38.

In the endocrine pancreas in vivo, treatment with GCs results in secondary β cell hyperplasia in an effort to generate sufficient insulin to maintain normoglycemia in the setting of the peripheral insulin resistance in various target tissues³⁰. However, studies with isolated islets in culture have demonstrated conflicting results. Direct treatment of isolated islets in culture with GCs induces β cell apoptosis³⁹. Glucose-stimulated insulin secretion (GSIS) depends on the influx of plasma glucose via the GLUT2 transporter present on β cells as well as glucokinase activity to generate glucose-6-phosphate (G6P) for entry into the glycolytic cycle. This process results in an increase in the ATP/ADP ratio and drives membrane depolarization in response to closure of ATP-sensitive potassium channels, leading to calcium influx via voltage gated calcium channels and ultimately exocytosis of insulin-containing granules. Notably, GCs impair GSIS on multiple levels³⁰. GCs promote the degradation of GLUT2 and reduce expression levels of glucokinase^40,41. Additionally, the activity of glucose-6-phosphatase is increased, further impairing the entry of G6P into the glycolytic cycle⁴². GCs also enhance repolarization of the β cell by upregulation of the Kv1.5 voltage-gated potassium channel, thus attenuating calcium influx and exocytosis of insulin-containing granules⁴³. Indeed, in vivo overexpression of GR in the β cell to address the discrepancy between these in vivo and in vitro findings has demonstrated glucose intolerance and impaired insulin secretion in these transgenic mice^30,44.

Summary and Future Directions

Glucocorticoids play an intricate role in all aspects of glucose metabolism. Our review has focused on mechanisms of GC regulation and signaling, clinical and subclinical syndromes of GC excess, and the physiology of GC-regulated glucose homeostasis. While the utility of GC upregulation during a short-term stressful event remains clear, long-term GC excess, whether endogenous or exogenous, is wrought with complications. Overt clinical GC excess is observed in the setting of CS, while subclinical GC excess may be present in those with metabolic syndrome.

Given the associated increase in mortality and morbidity in these patients, studies are actively underway to better understand the utility of small molecules capable of modulating GC signaling. These agents are of significant interest clinically in the treatment of Cushing’s syndrome, insulin resistance/DM, and autoimmune/inflammatory conditions. The development of selective glucocorticoid receptor modulators (SEGRMs) capable of harnessing the anti-inflammatory properties of GCs while avoiding the long-term metabolic consequences remains an attractive concept but has proven difficult in practice because of the overlap that exists between GR transrepression and transactivation in the regulation of these two paradigms⁴⁵. More recently, mifepristone, a GR antagonist, was approved for the treatment of Cushing’s syndrome with glucose intolerance/DM in patients refractory to standard surgical management⁴⁶. However, the systemic use of a glucocorticoid antagonist in patients without glucocorticoid excess for the treatment of diabetes is likely to be associated with symptoms of adrenal insufficiency. Consistent with the dehydrogenase hypothesis, liver-specific GR antagonists are in development for the treatment of diabetes. One such compound is a GR antagonist fused to a liver targeting bile acid. Such compounds should reduce hepatic gluconeogenesis without affecting the HPA axis⁴⁷. Another promising approach is the generation of antisense oligonucleotides targeting GR mRNA specifically in the liver⁴⁸. Lastly, the use of 11βHSD1 inhibitors as a means of attenuating GC action at the tissue level continues to hold promise as these agents demonstrate measurable improvements in insulin resistance and blood glucose measurements as well as other indices associated with metabolic syndrome. However, the consequences of the resulting long-term upregulation of the HPA axis and immunomodulatory effects remain unclear⁴⁹.

In summary, the mechanisms underlying the relationship between glucocorticoids and diabetes are not fully understood. What remains clear is that GCs are capable of regulating aspects of glucose homeostasis in each target organ by antagonizing the effects of insulin either directly or indirectly. The marked interest in the modulation of GC signaling as a therapeutic tool for diabetes and metabolic syndrome will undoubtedly lead to a better understanding of these complex mechanisms in the years to come as our ability to dissect the complex tissue-specific effects of GCs in the laboratory improve.

Biography

Kevin T. Bauerle, MD, PhD, (left), is a Clinical Fellow, and Charles Harris, MD, PhD, (right), is an Assistant Professor, Division of Endocrinology, Metabolism, and Lipid Research, Department of Medicine, Washington University School of Medicine, St. Louis.

Contact: harrisc@wustl.edu