Essential role of brain-derived neurotrophic factor in the regulation of serotonin transmission in the basolateral amygdala

Abstract

Human and animal model studies have linked brain-derived neurotrophic factor (BDNF) with the etiology of anxiety disorders. This pleiotropic neurotrophin and its receptor, TrkB, promote neuronal survival, differentiation and synaptic plasticity. Here we interrogated the role of BDNF in serotonergic neurotransmission in the basolateral amygdala (BLA), a limbic brain region associated with the neurobiology of anxiety. We found that both GABAergic and pyramidal projection neurons in the wild-type BLA contained TrkB receptors. Examination of BDNF(2L/2LCk-Cre) mutant mice with brain-selective depletion of BDNF revealed mild decreases in serotonin content in the BLA. Notably, whole cell recordings in BLA pyramidal cells uncovered significant alterations in 5-HT(2)-mediated regulation of GABAergic and glutamatergic transmission in BDNF(2L/2LCk-Cre) mutant mice that result in a hyperexcitable circuit. These changes were associated with decreased expression of 5-HT(2) receptors. Collectively, the results indicate a required role of BDNF in serotonin transmission in the BLA. Furthermore, they suggest a mechanism underlying the reported increase in anxiety-like behavior elicited by perturbed BDNF signaling.

Figure 1. TrkB receptors are expressed in BLA Projection and GABAergic neurons

Representative images illustrating co-localization of TrkB receptor with CamK and GAD67 in the wild-type BLA (arrows). CamK marks pyramidal neurons and GAD denotes GABAergic interneurons. Scale bars = 10 μm

Figure 2. Depletion of BDNF mRNA in the BLA and raphe nucleus of BDNF^2L/2LCk-Cre mutant mice

In situ hybridization analysis of representative coronal brain sections obtained from wild-type (A and B) and BDNF^2L/2LCk-Cre mutant mice (C and D) showing extensive depletion of BDNF mRNA in the BLA (A and C) and dorsal raphe nucleus (B and D) (arrows) of BDNF^2L/2LCk-Cre mice compared to wild-types.

Figure 3. Serotonin content in the BLA of wild-type and BDNF^2L/2LCk-Cre mutant mice

Levels of (A) serotonin (B), 5-HIAA and (C) 5HIAA/5-HT ratio in the BLA of wild-type (WT) and BDNF^2L/2LCk-Cre (CM) mice. A significant decrease in 5-HT content was observed in BDNF^2L/2LCk-Cre mice whereas levels of 5-HIAA and the 5-HT/5-HIAA ratio were normal. *, p = 0.03; n = 8 animals.

Figure 4. 5-HT_2A-mediated inhibitory responses in BLA pyramidal neurons are reduced in BDNF^2L/2LCk-Cre mutant mice

Traces in voltage clamp from wild-type (A) and BDNF^2L/2LCk-Cre (B) mice show inhibitory postsynaptic currents (IPSCs) under basal conditions, in the presence of serotonin (5-HT; 50 μM, 1 min), 5-HT and bicuculline (Bic., 20 μM) or 5-HT and cinanserine (Cin., 20 μM). Serotonin elicits a significant increase in spontaneous IPSC frequency in neurons from wild-type (p = 0.03) but not mutant (p = n/s) mice. C, Bar graph summarizes IPSC frequency data (mean ± SE) for wild-type (WT) and BDNF^2L/2LCk-Cre mice (CM) following 5-HT application and in the presence of bicuculline (Bic), cinanserine (Cin.) and MDL 100,907 (MDL). *, p = 0.01; n = 6 (wild types) and 7 (BDNF^2L/2LCk-Cre) cells.

Figure 5. BLA pyramidal neurons in BDNF^2L/2LCk-Cre mutants show exaggerated 5-HT2-mediated excitatory responses

A, Representative traces in voltage clamp from wild-type and BDNF^2L/2LCk-Cre conditional mutant mice show excitatory postsynaptic currents (EPSCs) under basal conditions and in the presence of serotonin (5-HT, 50 μM, 1 min). B, Bar graph summarizes EPSC frequency data (mean ± SE) for wild-type (WT) and BDNF^2L/2LCk-Cre mutant (CM) mice following 5-HT application. This effect was blocked in the presence of NBQX (5-HT and NBQX) or cinanserine (5-HT and Cin.). The data are based on recordings from wild-type (n = 16) and BDNF^2L/2LCk-Cre mutant (n = 21) responding pyramidal cells. While serotonin elicits a significant increase in spontaneous EPSC frequency in BLA pyramidal neurons from both BDNF^2L/2LCk-Cre (p < 0.0001) and wild-type (p = 0.01) mice, the excitatory effect was more pronounced in mutant neurons and blocked by NBQX and cinanserine. *, p = 0.01; **, p = 0.03; ***, p = 0.004.

Figure 6. Expression of 5-HT_2A and 5-HT_2C receptor mRNA in the BLA

Content of 5-HT_2A (A) and 5-HT_2C (B) receptor mRNA in BLA tissue punches obtained from wild-type (WT) and BDNF^2L/2LCk-Cre (CM) mice (n = 6). Data are expressed as expression of 5-HT receptor mRNA in BDNF^2L/2LCk-Cre mice relative to wild-type controls. Diminished BDNF signaling resulted in decreased 5-HT_2A and 5-HT_2C receptor mRNA levels in the BLA. *, p < 0.01

Figure 7. Modulation of glutamatergic and GABAergic transmission by 5-HT and BDNF in the BLA

We propose that BDNF influences the balance of excitatory and inhibitory transmission in the BLA thereby affecting overall BLA output. Our data suggest that BDNF positively regulates expression of pre-synaptic G_q-coupled 5-HT₂ receptors on GABA interneurons, thereby mediating increases in GABAergic inhibitory drive onto glutamatergic pyramidal neurons induced by 5-HT released by dorsal raphe nucleus (DRN) fibers. As proposed previously (Rainnie 1999), serotonin might also act on 5-HT₂ receptors on glutamatergic inputs onto pyramidal neurons to positively regulate glutamate release and prevent prolonged inhibition. Our findings suggest that BDNF might also influence this balancing effect of 5-HT on glutamatergic transmission in the BLA.