Back-propagating action potential

Abstract

Brain derived neurotrophic factor (BDNF) is crucial for the formation of appropriate synaptic connections during development and for learning and memory in adults. Secretion of this neurotrophin is under activity-dependent control. Understanding which patterns of physiological activity regulate BDNF secretion is therefore an important step in the comprehension of its role. We have recently shown that back propagation of action potentials (bAPs) is the principal triggering mechanism of dendritic BDNF secretion occurring during ongoing neuronal activity in neuronal cultures. In the present addendum we discuss possible implications of bAPs-induced BDNF secretion on the construction and reorganization of neuronal networks.

Brain derived neurotrophic factor (BDNF) is the most abundantly expressed neurotrophin in the mammalian nervous system. BDNF is a small-secreted homodimeric polypeptide (monomeric MW: 14 kDa) that binds to two structurally unrelated membrane receptors: the high affinity tyrosine kinase receptors (TrkB) or the low affinity pan-neurotrophin binding receptor p75^NTR.¹ Once released into the extracellular space, BDNF binds to the TrkB receptor and regulates nearly all keys steps of network development from neuronal migration and differentiation to experience-dependent refinement of local connections.² BDNF participates in several forms of activity-dependent synaptic plasticity in both developing and adult central nervous system.³ The critical requirement of BDNF in brain function and development is illustrated by the observation that several neuropathological conditions in human are associated with deficits in BDNF secretion.^4–6 Given the crucial role of BDNF signaling in the functional and structural organization of both developing and adult nervous system a big effort has been initiated over the past decade to study the mode and sites of BDNF secretion.

In the CNS BDNF is synthesized by the principal, i.e., glutamatergic, neurons in which BDNF is packaged into secretory granules in both axon terminals and dendrites. Two modalities of BDNF secretion can exist in cultured neurons: a constitutive Ca²⁺-independent secretion that occurs in the absence of any specific triggering events and a Ca²⁺-dependent secretion that is regulated by synaptic activity.⁷ Recent studies using fluorescent-labeled BDNF have directly demonstrated that BDNF can be released from both axonal8 and dendritic processes.⁹ However, although these studies have led to important information on the mechanisms and site of BDNF release, the physiologically relevant patterns of electrical activity that trigger BDNF release could not be delineated.

In our recent paper,¹⁰ we combined electrophysiological recording, time-lapse fluorescence imaging and immuno-staining on BDNF-GFP expressing hippocampal neurons in culture to study how ongoing activity affects BDNF secretion from neuronal dendrites. Our results show that the generation of action potentials that back-propagate into the dendrites (bAPs) is the key signal for dendritic BDNF release during ongoing synaptic activity. This release is achieved through the activation of voltage-gated calcium channels and subsequent influx of Ca²⁺ into the intracellular space. As already reported, the release of BDNF is a relatively slow process (in the order of several seconds). As few as 4 bAPs are sufficient to induce dendritic BDNF secretion. Moreover the probability of b-APs to trigger BDNF secretion, but not the amount of BDNF secreted, was dependent on the number of bAPs, suggesting that bAPs acts to trigger the release, i.e., once the secretion threshold is reached additional firing activity did not further modify it. Interestingly, we also show that the spontaneous synaptic activity alone, in the absence of APs, is not sufficient to induce dendritic BDNF release (Fig. 1). This point is of particular interest because it suggests that dendritic BDNF release is a function of the neuronal output, i.e., APs, rather than its synaptic input. Therefore all mechanisms that regulate the generation of the APs, such as peri-somatic inhibition, or that affect the back propagation of the APs into the dendrites, such as the activation of the cholinergic and β-adrenergic systems,^11,12 or the shunting effect exerted by dendritic GABAergic conductance¹³ would be able to control dendritic BDNF secretion independently of the synaptic drive received by the neuron. Our results also suggest that in response to identical synaptic inputs, a neuron could undergo BDNF-dependent morphological and physiological reorganization depending on whether or not the environmental conditions allow APs to back propagate into the dendrites (Fig. 2). In that respect, it is interesting to note that suppressing AP generation in a single neuron within a network of active cultured neurons, by overexpressing an inward-rectifier potassium channel, affects the strength of the glutamatergic synapses impinging on that specific neuron.¹⁴ In light of our results and the known action of BDNF on synaptic plasticity it is plausible that such effect could be, at least in part, mediated by a reduction of dendritic BDNF secretion.

Figure 1.

b-APs trigger dendritic BDNF secretion. Neuron transfected with BDNF-GPF show a dendritic distribution of BDNF into secretory granules. BDNF-GFP release (decreased of granular fluorescence) is triggered by firing activity (right) but not by synaptic activity (left).

Figure 2.

BDNF secretion depends on the neuronal output rather than synaptic inputs. Supra-threshold excitatory postsynaptic potentials (EPSPs) trigger action potentials that back-propagate (b-APs) from the soma to the dendrites. Center: the b-APs induced BDNF secretion and synaptic strengthening. Left: this process can be reinforced by conditions that facilitate the back propagation of APs such as the neuromodulatory action of ACh. Right: when the same supra-threshold glutamatergic activity is generated in a context that prevents APs backpropagation, such as a concomitant activation of dendritic inhibition (GABA), BDNF secretion and synaptic strengthening will not occur.

Is the amount of endogenous BDNF released from a single neuron following b-APs sufficient to have a physiological effect? To answer this question, we have shown that AP firing triggered in one single neuron induced a BDNF-dependent phosphorylation of the cyclic AMP response element-binding protein (CREB) in the neighboring neurons.¹⁵ CREB phosphorylation represents one of the early steps in synaptic plasticity consolidation subtending memory formation.¹⁶ In light of the fact that the activity of the modulatory systems such as the cholinergic system is involved in synaptic plasticity as well as in mnemonic processes,¹⁷ our results suggest a possible means by which neuromodulator action could affect memory formation through facilitation of bAPs that, in turn, trigger dendritic release of BDNF and CREB activation (Fig. 3). The question how localized the secretion of BDNF induced by b-APs can be, and identifying the associated biological effects underlying synapse specific release of BDNF, will be challenging tasks for future investigations. Previous studies have reported a contribution of NMDA receptors in dendritic BDNF release induced by tetanic stimulation.⁹ Release confined to individual synapses could influence synapse development and plasticity locally. In contrast, because APs and Ca²⁺ rise travel backward from the soma and have the capacity to invade the dendritic tree, BDNF secretion under these circumstances could influence synapse development and strength globally. Alternatively presynaptic activity concomitant with BDNF secretion can still constitute the discriminating factor for plastic modifications as recently shown by Tanaka and collaborators.¹⁸

Figure 3.

Modulation of back propagating action potentials can affect memory consolidation through dendritic BDNF release and CREB phosphorylation.

Addendum to: Kuczewski N, Porcher C, Ferrand N, Fiorentino H, Pellegrino C, Kolarow R, Lessmann V, Medina I, Gaiarsa JL. Backpropagating action potentials trigger dendritic release of BDNF during spontaneous network activity. J Neurosci. 2008;28:7013–7023. doi: 10.1523/JNEUROSCI.1673-08.2008.