Type 1 diabetes, glucocorticoids and the brain: a sweet connection

Abstract

Peripheral and autonomous neuropathies are well-known and devastating complications of type 1 diabetes (T1D). However, T1D can also impact the integrity of the central nervous system (CNS), and the reason why T1D affects CNS integrity remains to be elucidated. Studies on diabetic patients demonstrated mild to moderate slowing of mental speed and diminished mental flexibility. Although the alterations in cognitive functions under normal conditions are not severe, mild cognitive defects can influence everyday activities in more demanding situations. Indeed, in 2004 Sandeep et al reported that hypercortisolism in diabetic patients may contribute to their hippocampal dysfunction. To investigate disease initiation, progression, and treatments without exposing humans to unnecessary and potentially unethical risks, animal models have been developed. The physiology of mice, rats, and other animals is remarkably conserved in comparison to the human condition, and over the last 40 years several animal models have become available. We have used two animal models, i.e. a pharmacological model, the streptozotocin (STZ) treated mouse, and a genetic model the NOD mouse, which spontaneously develops type 1 diabetes. As type 1 diabetic patients, these animal models show high circulating glucocorticoid levels, increased sensitivity to stress, and morphological alteration in various brain areas. In the present study these models were used to test the hypothesis that the onset of diabetes induces first dysregulation of the hypothalamus-pituitary-adrenal (HPA) axis and subsequently hypersecretion of glucocorticoids which then renders the brain more vulnerable to metabolic insults causing damage and concomitant cognitive disturbances. The NOD model revealed a surge in ACTH release, which likely preceded the onset and progression of diabetes marked by adrenal hyperreponsiveness and hypersecretion of corticosterone. To our surprise we found in the STZ model that not the initial ACTH surge was the most proximal cause of hypercorticism in diabetes, but rather the induction of adrenocortical ACTH receptors per se. At no time-point after STZ administration ACTH levels did rise reinforcing the notion that hyperresponsiveness of the adrenals to ACTH may occur independent of the mitogenic activity of the peptide. In the same model, excess glucocorticoids rather than glycemia and insulin appeared causal to cerebral damage and mild cognitive impairment. These deficits in hippocampal function induced by high glucocorticoid concentrations were readily ameliorated by a brief treatment with the glucocorticoid receptor antagonist mifepristone. These findings make the GRs a suitable target for new therapeutic strategies aimed to normalize the disturbed hippocampal functions characteristic for diabetes neuropathology.