Role of the 5-HT4 receptor in chronic fluoxetine treatment-induced neurogenic activity and granule cell dematuration in the dentate gyrus

Abstract

Background: Chronic treatment with selective serotonin (5-HT) reuptake inhibitors (SSRIs) facilitates adult neurogenesis and reverses the state of maturation in mature granule cells (GCs) in the dentate gyrus (DG) of the hippocampus. Recent studies have suggested that the 5-HT4 receptor is involved in both effects. However, it is largely unknown how the 5-HT4 receptor mediates neurogenic effects in the DG and, how the neurogenic and dematuration effects of SSRIs interact with each other.

Results: We addressed these issues using 5-HT4 receptor knockout (5-HT4R KO) mice. Expression of the 5-HT4 receptor was detected in mature GCs but not in neuronal progenitors of the DG. We found that chronic treatment with the SSRI fluoxetine significantly increased cell proliferation and the number of doublecortin-positive cells in the DG of wild-type mice, but not in 5-HT4R KO mice. We then examined the correlation between the increased neurogenesis and the dematuration of GCs. As reported previously, reduced expression of calbindin in the DG, as an index of dematuration, by chronic fluoxetine treatment was observed in wild-type mice but not in 5-HT4R KO mice. The proliferative effect of fluoxetine was inversely correlated with the expression level of calbindin in the DG. The expression of neurogenic factors in the DG, such as brain derived neurotrophic factor (Bdnf), was also associated with the progression of dematuration. These results indicate that the neurogenic effects of fluoxetine in the DG are closely associated with the progression of dematuration of GCs. In contrast, the DG in which neurogenesis was impaired by irradiation still showed significant reduction of calbindin expression by chronic fluoxetine treatment, suggesting that dematuration of GCs by fluoxetine does not require adult neurogenesis in the DG.

Conclusions: We demonstrated that the 5-HT4 receptor plays an important role in fluoxetine-induced adult neurogenesis in the DG in addition to GC dematuration, and that these phenomena are closely associated. Our results suggest that 5-HT4 receptor-mediated phenotypic changes, including dematuration in mature GCs, underlie the neurogenic effect of SSRIs in the DG, providing new insight into the cellular mechanisms of the neurogenic actions of SSRIs in the hippocampus.

Figure 1

5-HT levels in the hippocampus of 5-HT4R KO mice with the C57BL/6 J-background. (A) Tissue 5-HT and 5-HIAA levels in the hippocampus and raphe nuclei. Data are expressed as the mean ± SEM (n = 7 or 8). N.S., not significant for unpaired t-tests. (B) Fluoxetine (Flx)-induced increase in extracellular 5-HT concentrations in the hippocampus. After equilibration, dialysate samples were collected every 20 min with either fluoxetine (22 mg/kg) or saline (Sal) administered at time zero. The average value of seven basal samples for each animal was defined as 100% and used for normalization. Data are expressed as the mean ± SEM (n = 4 for fluoxetine, n = 3 for saline). Main effect of group: P = 0.0499; main effect of time point: P < 0.0001; interaction of group and time point: P < 0.0001. P value was determined by two-way ANOVA for repeated measures. * P < 0.01 [WT(Sal) vs WT(Flx)] for post hoc Bonferroni’s test after two-way ANOVA; N.S, not significant [WT(Flx) vs KO(Flx)] for post hoc Bonferroni’s test at every time point after two-way ANOVA. (C) Tissue 5-HT and 5-HIAA levels in the hippocampus after chronic fluoxetine treatment. Fluoxetine (22 mg/kg) was administered for 3 weeks. Data are expressed as the mean ± SEM (n = 7 or 8). Main effect of drug: ^#P < 0.0001 (5-HT), ^#P < 0.0001 (5-HIAA); main effect of genotype: P = 0.9734 (5-HT), P = 0.0423 (5-HIAA); interaction of drug and genotype: P = 0.9790 (5-HT), P = 0.3243 (5-HIAA); P values determined by two-way ANOVA.

Figure 2

Effect of chronic fluoxetine treatment on adult neurogenesis in 5-HT4R KO mice. (A) Experimental scheme. Mice were intraperitoneally (ip) injected with fluoxetine (Flx) at a dose of 22 mg/kg for 21 days and were administered BrdU 24 h after the last treatment (on day-22) at a dose of 150 mg/kg. Mice were sacrificed 2 h after the BrdU injection. (B) Immunohistochemical visualization of BrdU in the SGZ of the DG. Scale bar: 100 μm. Arrows represent BrdU-positive cells. (C) Quantification of BrdU-positive cells in the SGZ of the DG in WT mice and 5-HT4R KO mice. Data are expressed as the mean ± SEM (n = 4 or 5). Main effect of drug: P = 0.0023; main effect of genotype: P = 0.0107; interaction of drug and genotype: P = 0.0048; P values determined by two-way ANOVA. *** P < 0.001 and N.S, not significant for post hoc Bonferroni’s test, respectively, after two-way ANOVA. (D) Representative images of anti-doublecortin (DCX) immunohistochemistry in the DG. Scale bar: 100 μm. (E) Quantification of DCX-positive immature neurons in the DG of the WT mice and 5-HT4R KO mice. The number of DCX-positive cells in the DG is shown as the number of cells per square millimeter of DG area. Data are expressed as the mean ± SEM (n = 4 or 5). Main effect of drug: P = 0.0214; main effect of genotype: P = 0.1372; interaction of drug and genotype: P = 0.0032; P values determined by two-way ANOVA. ** P < 0.01 and N.S., not significant for post hoc Bonferroni’s test, respectively, after two-way ANOVA.

Figure 3

Expression of the 5-HT₄ receptor in the DG. (A) in situ hybridization analysis of 5-HT₄ receptor mRNA expression. GCL: granular cell layer, SGZ: subgranular zone. Scale bars: 100 μm (left) and 30 μm (right). (B) Representative images of β-galactosidase (LacZ) and markers of granule cells; NeuN, calbindin, doublecortin (DCX), and calretinin. Scale bar: 20 μm. Arrows represent DCX-positive cells. Arrow heads represent calretinin-positive cells.

Figure 4

Correlation between enhancement of adult neurogenesis and dematuration of GCs. (A) Representative images of calbindin-IR. GCL: granule cell layer, ML: molecular layer. Scale bar: 100 μm. (B) Quantification of calbindin-IR in the GCL and the ML in WT mice and 5-HT4R KO mice. Data are expressed as the mean ± SEM (n = 4 or 5). Main effect of drug: P = 0.0014 (GCL), P = 0.0015 (ML); main effect of genotype: P = 0.0237 (GCL), P = 0.0208 (ML); interaction of drug and genotype: P = 0.0022 (GCL), P = 0.0009 (ML); P values determined by two-way ANOVA. *** P < 0.001 and N.S., not significant for post hoc Bonferroni’s test, respectively, after two-way ANOVA. (C) Comparison between the number of BrdU-positive cells in the SGZ and calbindin-IR intensity of the granule cell layer in WT mice. Mice that received fluoxetine treatment (22 mg/kg) for 3 or 4 weeks were plotted in the analysis. The Pearson correlation coefficient (R) was calculated (P = 0.0278). (D) Comparison of gene expression changes after chronic fluoxetine treatment between calbindin (Calb1) and Bdnf or Npy. Gene expression was normalized by the average gene expression in control mice. Mice received fluoxetine treatment (22 mg/kg) for 3 or 4 weeks. Four independent experiments were included in the analysis. The Pearson correlation coefficient (R) was calculated (Bdnf vs. Calb1, P = 0.0002, Npy vs. Calb1, P < 0.0001).

Figure 5

Effect of X-ray irradiation on fluoxetine-induced adult neurogenesis and dematuration in the DG. (A) Experimental scheme. Mice were irradiated with X-rays (10 Gy) 14 days before initiation of fluoxetine treatment (ip, intraperitoneally injected). Mice were sacrificed 24 h after the final drug treatment. (B and C) Representative images of DCX (B) and calbindin (C) immunohistochemistry. Scale bars: 100 μm. (D) Quantification of calbindin-IR signal intensity in the GCL and the ML in control (Sham) and X-ray-irradiated mice. Data are expressed as the mean ± SEM (n =5 or 6). Main effect of drug: P < 0.0001 (GCL), P < 0.0001 (ML); main effect of X-ray irradiation: P = 0.0303 (GCL), P = 0.0087 (ML); interaction of drug and X-ray irradiation: P = 0.1089 (GCL), P = 0.2085 (ML); P values determined by two-way ANOVA.

Figure 6

Model of 5-HT₄ receptor-mediated increased neurogenesis and GC dematuration in the DG induced by chronic SSRIs. Chronic SSRI treatment induces dematuration of GCs and increases adult neurogensis via the 5-HT₄ receptor (5-HT₄R) in the DG. Dematured GCs exhibit an increase in expression of neurogenic factors, which may be involved in enhanced adult neurogenesis in the SGZ of the DG.