Evidence for preserved insulin responsiveness in the aging rat brain

Abstract

Abstract Insulin appears to exert salutary effects in the central nervous system (CNS). Thus, brain insulin resistance has been proposed to play a role in brain aging and dementia but is conceptually complex and unlikely to fit classic definitions established in peripheral tissues. Thus, we sought to characterize brain insulin responsiveness in young (4–5 months) and old (24 months) FBN male rats using a diverse set of assays to determine the extent to which insulin effects in the CNS are impaired with age. When performing hyperinsulinemic-euglycemic clamps in rats, intracerebroventricular (ICV) infusion of insulin in old animals improved peripheral insulin sensitivity by nearly two-fold over old controls and comparable to young rats, suggesting preservation of this insulin-triggered response in aging per se (p < 0.05). We next used an imaging-based approach by comparing ICV vehicle versus insulin and performed resting state functional magnetic resonance imaging (rs-fMRI) to evaluate age- and insulin-related changes in network connectivity within the default mode network. In aging, lower connectivity between the mesial temporal (MT) region and other areas, as well as reduced MT signal complexity, was observed in old rats, which correlated with greater cognitive deficits in old. Despite these stark differences, ICV insulin failed to elicit any significant alteration to the BOLD signal in young rats, while a significant deviation of the BOLD signal was observed in older animals, characterized by augmentation in regions of the septal nucleus and hypothalamus, and reduction in thalamus and nucleus accumbens. In contrast, ex vivo stimulation of hippocampus with 10 nM insulin revealed increased Akt activation in young (p < 0.05), but not old rats. Despite similar circulating levels of insulin and IGF-1, cerebrospinal fluid concentrations of these ligands were reduced with age. Thus, these data highlight the complexity of capturing brain insulin action and demonstrate preserved or heightened brain responses to insulin with age, despite dampened canonical signaling, thereby suggesting impaired CNS input of these ligands may be a feature of reduced brain insulin action, providing further rationale for CNS replacement strategies.