Reduced hippocampal neurogenesis and number of hilar neurones in streptozotocin-induced diabetic mice: reversion by antidepressant treatment.

Cerebral dysfunctions, including a high incidence of depression, are common findings in human type 1 diabetes mellitus. An association between depression and defective hippocampal neurogenesis has been proposed and, in rodents, antidepressant therapy restores neuronal proliferation in the dentate gyrus. Hippocampal neurogenesis is also deficient in diabetic mice, which led us to study whether the selective serotonin reuptake inhibitor fluoxetine influences cell proliferation in streptozotocin-diabetic animals. Diabetic and control C57BL/6 mice received fluoxetine (10 mg/kg/day, i.p., 10 days) and dentate gyrus cell proliferation was measured after a single injection of 5-bromo-2'-deoxyuridine (BrdU). Diabetic mice showed reduced cell proliferation. Fluoxetine treatment, although having no effect in controls, corrected this parameter in diabetic mice. The phenotype of newly generated cells was analysed by confocal microscopy after seven daily BrdU injections, using Tuj-1/beta-III tubulin as a marker for immature neurones and glial fibrillary acidic protein for astrocytes. In controls, the proportion of Tuj-1-BrdU-positive cells over total BrdU cells was approximately 70%. In vehicle-treated diabetic mice, immature neurones decreased to 56% and fluoxetine brought this proportion back to control values without affecting astrocytes. Therefore, fluoxetine preferentially increased the proliferation of cells with a neuronal phenotype. In addition, neurones were counted in the hilus of the dentate gyrus; a 30% decrease was found in diabetic mice compared with controls, whereas this neuronal loss was prevented by fluoxetine. In conclusion, fluoxetine treatment restored neuroplasticity-related hippocampal alterations of diabetic mice. These findings may be potentially important to counteract diabetes-associated depression in humans.