Physical exercise and antidepressants enhance BDNF targeting in hippocampal CA3 dendrites: further evidence of a spatial code for BDNF splice variants

Abstract

Brain-derived neurotrophic factor (BDNF) is encoded by multiple BDNF transcripts, whose function is unclear. We recently showed that a subset of BDNF transcripts can traffic into distal dendrites in response to electrical activity, while others are segregated into the somatoproximal domains. Physical exercise and antidepressant treatments exert their beneficial effects through upregulation of BDNF, which is required to support survival and differentiation of newborn dentate gyrus (DG) neurons. While these DG processes are required for the antidepressant effect, a role for CA1 in antidepressant action has been excluded, and the effect on CA3 neurons remains unclear. Here, we show for the first time that physical exercise and antidepressants induce local increase of BDNF in CA3. Voluntary physical exercise for 28 consecutive days, or 2-week treatment with 10 mg/kg per day fluoxetine or reboxetine, produced a global increase of BDNF mRNA and protein in the neuronal somata of the whole hippocampus and a specific increase of BDNF in dendrites of CA3 neurons. This increase was accounted for by BDNF exon 6 variant. In cultured hippocampal neurons, application of serotonin or norepinephrine (10-50 μM) induced increase in synaptic transmission and targeting of BDNF mRNA in dendrites. The increased expression of BDNF in CA3 dendrites following antidepressants or exercise further supports the neurotrophin hypothesis of antidepressants action and confirms that the differential subcellular localization of BDNF mRNA splice variants provides a spatial code for a selective expression of BDNF in specific subcellular districts. This selective expression may be exploited to design more specific antidepressants.

Figure 1

Effects of physical exercise on local brain-derived neurotrophic factor (BDNF) mRNA expression in mouse hippocampus. (a) High-magnification pictures of the in situ hybridization for BDNF mRNA in mouse hippocampal CA1, dentate gyrus (DG), and CA3 regions showing staining in the stratum pyramidale (CA1, CA3) and granularis (DG). Labeled dendrites are visible only in the CA3 stratum lucidum and radiatum (empty arrows). (b–d) Line graphs show densitometric analysis of the laminas containing dendrites, expressed as percentage of the staining measured in control animals (=100%). The distance is expressed in μm from the point of emergence of the apical dendrites from the soma (=0 μm). (e–g) Bar graphs show pooled data calculated by grouping intervals of 20 μm from the same hippocampal subregion. Each bar represents the mean±SE of six animals per group. ^*p<0.05. s.pyr, stratum pyramidale; p.rad, proximal radiatum; d.rad, distal radiatum; s.gr, stratum granularis; IML, inner molecular layer; MML, medial molecular layer; OML, outer molecular layer; s.luc, stratum lucidum; and s.rad, stratum radiatum.

Figure 2

Effects of antidepressants on local brain-derived neurotrophic factor (BDNF) mRNA expression in rat hippocampus. (a) Representative images of in situ hybridizations for total BDNF mRNA in dorsal rat hippocampus sections from animals treated with vehicle (CTRL), 10 mg/kg fluoxetine (FLX), or 10 mg/kg reboxetine (RBX). (b) High-magnification pictures from CA1, dentate gyrus (DG), and CA3 areas. Arrows indicate labeling of dendritic field in CA3. (c–e) Densitometric analysis of the dendritic labeling for total BDNF mRNA in CA1, DG, and CA3 expressed as gray level, a function of the distance from the cell soma (in μm) in control fluoxetine- and reboxetine-treated animals. Data are the means±SE of 4–6 animals; ^**p<0.01, ^***p<0.001. Neuroanatomical subregions are indicated as in Figure 1.

Figure 3

Effects of exercise or antidepressants on local brain-derived neurotrophic factor (BDNF) protein expression in rat hippocampus. (a, b) Anti-BDNF immunofluorescence (green, with Hoechst-counterstained nuclei in blue) in the hippocampus of animals from (a) sedentary, exercise, or lock-wheel groups; or (b) treated with vehicle (Ctrl), 10 mg/kg antidepressant fluoxetine (FLX), or reboxetine (RBX). Right panels show enlarged CA3 fields from the same sections. Arrowheads in the CA3 region in (a) and (b) point to increased BDNF immunofluorescence in the stratum lucidum and radiatum. (c, d) Densitometric quantification of BDNF in CA3 subregions of animals in the different conditions (n=4, each group) in comparison with BDNF fluorescence staining quantified in CA3 from control group. The differential staining visible in the figure does not affect the results in the densitometry because the different treatments were normalized on their internal controls. Data are the means±SD of four sections; ^*p<0.05. s.pyr, stratum pyramidale; p.rad, proximal radiatum; d.rad, distal radiatum.

Figure 4

Effects of exercise on exons 4 and 6 brain-derived neurotrophic factor (BDNF) mRNA expression in rat hippocampus. (a, b) High magnification of CA3 area in coronal brain sections at the level of the dorsal hippocampus showing localization of (a) exon 4 BDNF transcripts or (b) exon 6 BDNF transcripts in control sedentary mice, or accessing a lock-wheel or doing exercise on a running wheel. (c–h) Densitometric analysis of the dendritic labeling of exon 4 BDNF transcripts (c, e, g) or exon 6 BDNF (d, f, h) in CA1, dentate gyrus (DG), and CA3 expressed as gray level, a function of the distance from the cell soma (in μm). Data are the means±SE of 3–4 animals, ^*p<0.05; ^**p<0.01. Neuroanatomical subregions are indicated as in Figure 1. Scale bar in (a)=500 μm and in (b)=100 μm.

Figure 5

Effects of antidepressants on exons 4 and 6 brain-derived neurotrophic factor (BDNF) mRNA expression in rat hippocampus. (a, b) High magnification of CA3 area in coronal brain sections at the level of the dorsal hippocampus showing localization of (a) exon 4 BDNF transcripts or (b) exon 6 BDNF transcripts in control condition (CTRL) or following 10 mg/kg fluoxetine (FLX) or reboxetine (RBX) treatments. (c–h) Densitometric analysis of the dendritic labeling of exon 4 BDNF transcripts (c, e, g) or exon 6 BDNF (d, f, h) in CA1, DG, and CA3 expressed as gray level, a function of the distance from the cell soma (in μm). Data are the means±SE of 4–6 animals; ^**p<0.01; ^***p<0.001. Neuroanatomical subregions are indicated as in Figure 1. Scale bar in (a)=500 μm and in (b)=100 μm.

Figure 6

Analysis of serotonin (5-HT) and norepinephrine (NE) effects on brain-derived neurotrophic factor (BDNF) mRNA localization and spontaneous synaptic electrical activity in cultured rat hippocampal neurons. (a) Representative pictures of in situ hybridization for total BDNF mRNAs from nine div rat-cultured hippocampal neurons treated with increasing concentrations of 5-HT or NE. The images show the first 100 μm of apical dendrites whose pictures were linearized using the ImageJ-straighten plugin. (b) Densitometric analysis of the pictures shown in (a) at the indicated distances (expressed in μm) from soma, showing the point at which the in situ hybridization signal reached the background level (ie at densitometric arbitrary units (AU)=35) that was used to calculate the maximal distance of dendritic labeling (MDDL) in each case. (c) MDDL for total BDNF mRNA in hippocampal neuronal cultures treated as indicated. (d) MDDL for exon 6 BDNF mRNA variants in hippocampal neuronal cultures treated as indicated. (e) Examples of spontaneous synaptic currents in nine div-cultured neurons recorded under whole-cell patch-clamp technique at a holding potential of −60 mV, in control conditions and after addition of 5-HT or NE (two separate cells). (f) Quantitative analysis of the mean frequency (% of control) of spontaneous currents in hippocampal cells in control (=100%) and after addition of 5-HT (n=4) and NE (n=5). Note that both drugs significantly increased the spontaneous current frequency relative to control: ^*p<0.05; ^**p<0.01; and ^***p<0.001.

Figure 7

Revised model of the neurotrophic effects of antidepressant drugs. Brain-derived neurotrophic factor (BDNF) released from dentate gyrus (DG) granule cells is shown in red, while BDNF released from CA3 neurons is shown in blue. (1.BDNF) Exercise and antidepressants stimulate production and secretion of BDNF from DG granule cells providing local trophic support for survival and differentiation of DG subgranular neural stem cells and (2.BDNF) dendrites. (3.BDNF) Local production and release of BDNF from apical dendrites of CA3 neurons may provide the target-derived trophic support to promote innervation of CA3 from newly formed mossy fibers. (4.BDNF) At the same time, anterogradely transported BDNF in mossy fibers supports the further dendritic targeting of BDNF mRNA in CA3 neurons.