ERK1/2 activation is necessary for BDNF to increase dendritic spine density in hippocampal CA1 pyramidal neurons

Abstract

Brain-derived neurotrophic factor (BDNF) is a potent modulator of synaptic transmission and plasticity in the CNS, acting both pre- and postsynaptically. We demonstrated recently that BDNF/TrkB signaling increases dendritic spine density in hippocampal CA1 pyramidal neurons. Here, we tested whether activation of the prominent ERK (MAPK) signaling pathway was responsible for BDNF's effects on spine growth. Slice cultures were transfected with enhanced yellow fluorescent protein (eYFP) by particle-mediated gene transfer, and CA1 pyramidal neurons were imaged by laser-scanning confocal microscopy. We confirmed that BDNF (24 h) increases spine density in apical dendrites of CA1 neurons. The MEK (ERK kinase) inhibitors PD98059 and U0126 completely prevented the increase in spine density induced by BDNF, without having an effect on spine density by themselves. In contrast to its actions on cortical pyramidal neurons, BDNF had minor and rather localized effects on dendritic complexity in hippocampal pyramidal neurons, increasing the total length, but not the branching of apical dendrites within CA1 stratum radiatum, without affecting basal dendrites in stratum oriens. Our results support the hypothesis that the ERK-signaling pathway not only mediates long-term synaptic plasticity and hippocampal-dependent learning, but it is also involved in the structural remodeling of excitatory spine synapses triggered by neurotrophins.

Figure 1

BDNF increases spine density in eYFP-transfected CA1 pyramidal neurons. (A) NeuN immunocytochemistry allowed identification of all hippocampal subfields, CA1, CA3, and dentate gyrus (DG). White square indicates the region from which pyramidal neurons were selected for confocal imaging. (B) Confocal image of a representative eYFP-transfected CA1 pyramidal neuron from a serum-free slice. (C) Confocal image of a representative eYFP-transfected CA1 neuron (green), filled with Alexa-594 during whole-cell recording (red). (D) Representative continuous records of spontaneous excitatory synaptic currents. (E) Action potential train (top trace) evoked by somatic current injection (bottom trace, 200 pA, 0.8 sec). (F) Higher magnification views of representative segments of apical dendrites from control serum-free (SF) and BDNF-treated slices (scale bar, 2 μm). (G) Quantification of spine density, expressed per 10 μm of apical dendrite. (#) P < 0.001, Student t test, t = 4.615, n = 10-13.

Figure 2

ERK activity is necessary for BDNF to increase spine density in CA1 pyramidal neurons. (A) High magnification views of representative segments of apical dendrites (scale bar, 2 μm). Images (left to right) are as follows: serum-free (SF); BDNF; BDNF + PD98059 at 10 μM (BDNF + PD 10 μM); BDNF + PD98059 at 25 μM (BDNF + PD 25 μM); BDNF + U0126 at 10 μM (BDNF + U 10 μM); BDNF + U0126 at 25 μM (BDNF + U 25 μM); PD98059 at 25 μM alone (PD 25 μM); and U0126 at 25 μM alone (U 25 μM). (B) Quantification of spine density, expressed per 10 μm of apical dendrite. (#) P < 0.01 vs. SF group; (^*) P < 0.05 vs. BDNF-treated group; ANOVA followed by Newman Keul test, F(7, 64) = 3.475; n = 5-13. (C) Representative Western blots using antiphosphorylated ERK1 and ERK2 (P-ERK1 and P-ERK2) and anti-ERK1 and ERK2 (Total ERK1 and Total ERK2) antibodies in whole homogenates of hippocampal slice cultures. Lanes (left to right) are as follows: serum-free (SF); BDNF; BDNF + PD98059 at 10 μM (BDNF + PD 10 μM); BDNF + PD98059 at 25 μM (BDNF + PD 25 μM); BDNF + U0126 at 10 μM (BDNF + U 10 μM); BDNF + U0126 at 25 μM (BDNF + U 25 μM); PD98059 at 25 μM alone (PD 25 μM); and U0126 at 25 μM alone (U 25 μM).

Figure 3

BDNF does not affect the length or branching of either apical or basal dendrites of CA1 pyramidal neurons. Representative low magnification views of eYFP-transfected CA1 neurons, and full tracings from serum-free (A,C), and BDNF-treated slices (B,D; scale bar, 50 μm). Total length of apical (E) and basal (F) dendrites, and total number of apical (G) and basal (H) branch points. P < 0.05 in all cases; Student's t test, n = 7-9.

Figure 4

BDNF has small and localized effects on apical, not basal, dendritic length in CA1 pyramidal neurons. Total length of apical (A) and basal (B) dendrites as a function of distance from the soma. (^*) P < 0.05; Student's t test, n = 7-9.