Neuroprotection provided by dietary restriction in rats is further enhanced by reducing glucocortocoids.

Glucocorticoids (GC)--corticosterone (CORT) in rodents and cortisol in primates--are stress-induced hormones secreted by adrenal glands that interact with the hypothalamic pituitary axis. High levels of cortisol in humans are observed in neurodegenerative diseases, including Alzheimer's disease (AD) and Parkinson's disease (PD), as well as in diabetes, post-traumatic stress syndrome, and major depression. Experimental models of diabetes in rats and mice have demonstrated that reduction of CORT reduces learning and memory deficits and attenuates loss of neuronal viability and plasticity. In contrast to the negative associations of elevated GC levels, CORT is moderately elevated in dietary restriction (DR) paradigms which are associated with many healthy anti-aging effects including neuroprotection. We demonstrate here in rats that ablating CORT by adrenalectomy (ADX) with replenishment to relatively low levels (30% below that of controls) prior to the onset of a DR regimen (ADX-DR) followed by central administration of the neurotoxin, kainic acid (KA), significantly attenuates learning deficits in a 14-unit T-maze task. The performance of the ADX-DR KA group did not differ from a control group (CON) that did not receive KA and was fed ad libitum (AL). By contrast, the sham-operated DR (SHAM-DR KA) group, SHAM-AL KA group, and ADX-AL KA group demonstrated poorer learning behavior in this task compared to the CON group. Stereological analysis revealed equivalent DR-induced neuroprotection in the SH-DR KA and ADX-DR KA groups, as measured by cell loss in the CA2/CA3 region of the hippocampus, while substantial cell loss was observed in SH-AL and ADX-AL rats. A separate set of experiments was conducted with similar dietary and surgical treatment conditions but without KA administration to examine markers of neurotrophic activity, brain-derived neurotrophic factor (BDNF), transcriptions factors (pCREB), and chaperone proteins (HSP-70). Under these conditions, we noted elevations in both BDNF and pCREB in ADX DR rats compared to the other groups; whereas, HSP-70, was equivalently elevated in ADX-DR and SH-DR groups and was higher than observed in both SH-AL and ADX-AL groups. These results support findings that DR protects hippocampal neurons against KA-induced cellular insult. However, this neuroprotective effect was further enhanced in rats with a lower-than control level of CORT resulting from ADX and maintained by exogenous CORT supplementation. Our results then suggest that DR-induced physiological elevation of GC may have negative functional consequences to DR-induced beneficial effects. These negative effects, however, can be compensated by other DR-produced cellular and molecular protective mechanisms.