Corticosterone facilitates fluoxetine-induced neuronal plasticity in the hippocampus

Abstract

The hippocampal dentate gyrus has been implicated in a neuronal basis of antidepressant action. We have recently shown a distinct form of neuronal plasticity induced by the serotonergic antidepressant fluoxetine, that is, a reversal of maturation of the dentate granule cells in adult mice. This "dematuration" is induced in a large population of dentate neurons and maintained for at least one month after withdrawal of fluoxetine, suggesting long-lasting strong influence of dematuration on brain functioning. However, reliable induction of dematuration required doses of fluoxetine higher than suggested optimal doses for mice (10 to 18 mg/kg/day), which casts doubt on the clinical relevance of this effect. Since our previous studies were performed in naive mice, in the present study, we reexamined effects of fluoxetine using mice treated with chronic corticosterone that model neuroendocrine pathophysiology associated with depression. In corticosterone-treated mice, fluoxetine at 10 mg/kg/day downregulated expression of mature granule cell markers and attenuated strong frequency facilitation at the synapse formed by the granule cell axon mossy fiber, suggesting the induction of granule cell dematuration. In addition, fluoxetine caused marked enhancement of dopaminergic modulation at the mossy fiber synapse. In vehicle-treated mice, however, fluoxetine at this dose had no significant effects. The plasma level of fluoxetine was comparable to that in patients taking chronic fluoxetine, and corticosterone did not affect it. These results indicate that corticosterone facilitates fluoxetine-induced plastic changes in the dentate granule cells. Our finding may provide insight into neuronal mechanisms underlying enhanced responsiveness to antidepressant medication in certain pathological conditions.

Figure 1. Schematic diagram showing timeline of corticosterone and fluoxetine administration.

In control mice (CNT), corticosterone (CORT) or vehicle (VEH) was administered for 7 week. In fluoxetine-treated mice (FLX), fluoxetine was added during the last 4 weeks.

Figure 2. Corticosterone facilitates effects of fluoxetine on frequency facilitation.

(A) The time course of frequency facilitation induced by 1-Hz stimulation. Sample traces show averages of 15 consecutive fEPSPs during baseline and 1 Hz stimulation. Scale bar: 10 ms, 0.5 mV. (B) Pooled data showing facilitated effects of fluoxetine on frequency facilitation at 1 Hz (CORT effect: P<0.0001, FLX effect: P = 0.0008, CORT×FLX: P = 0.0062, n = 6 to 7) and 0.2 Hz (CORT effect: P<0.0001, FLX effect: P<0.0001, CORT×FLX: P = 0.0361, n = 5 to 7) in corticosterone-treated mice. ***P<0.001 compared with CNT/CORT. (C) Lack of changes in ratios of fEPSP to presynaptic fiber volley (FV) amplitude (n = 6 to 7). (D) Reduced synaptic facilitation induced by paired stimulation at 50-ms interval in corticosterone-treated mice (n = 6 to 7). CORT effect: ***P = 0.0001. Sample traces are from CNT/VEH and CNT/CORT groups. Scale bar: 10 ms, 0.2 mV.

Figure 3. Effects of fluoxetine on monoaminergic synaptic modulation in corticosterone-treated mice.

(A) Effects of fluoxetine on serotonin-induced synaptic potentiation. CORT effect: P = 0.0049, CORT×FLX: P = 0.0489 (n = 5 to 7). (B) Facilitated effects of fluoxetine on synaptic potentiation induced by dopamine (10 µM). CORT effect: P = 0.0006, FLX effect: P = 0.0004, CORT×FLX: P = 0.0336 (n = 6 to 7). ***P<0.001 compared with CNT/CORT. The bar graph at right includes the results from slices pretreated with SCH23390 (n = 4 slices each). (C) Facilitated effects of fluoxetine on synaptic potentiation induced by SKF81297 (100 nM). CORT effect: P<0.0001, FLX effect: P = 0.0004, CORT×FLX: P = 0.0011 (n = 3 to 4 slices). ***P<0.001 compared with CNT/CORT.

Figure 4. Corticosterone facilitates effects of fluoxetine on expression of mature granule cell markers.

Calbindin, FLX effect: P = 0.0009, CORT×FLX: P = 0.0257. Desmoplakin, FLX effect: P = 0.0186. TDO, CORT effect: P = 0.011, FLX effect: P = 0.0086 (n = 4 each). *P<0.05, **P<0.01, ***P<0.001 compared with CNT/CORT.