A model of spike initiation in neocortical pyramidal neurons

Local and propagated dendritic action potentials evoked by glutamate iontophoresis on rat neocortical pyramidal neurons

Iontophoresis of glutamate at sites on the apical dendrite 278-555 microm from the somata of rat neocortical pyramidal neurons evoked low-threshold, small, slow spikes and/or large, fast spikes in 71% of recorded cells. The amplitude of the small, slow spikes recorded at the soma averaged 9.1 mV, and their apparent threshold was <10 mV positive to resting potential. Both their amplitude and their apparent threshold decreased as the iontophoretic site was moved farther from the soma. These spikes were not abolished by somatic hyperpolarization. When the somata of cells displaying these small spikes were voltage clamped at membrane potentials that prevented somatic or axonic firing, corresponding current spikes could be evoked all-or-none by dendritic depolarization, indicating that the small, slow spikes arose in the dendrite. Similar responses were not observed during somatic depolarization evoked by current pulses or glutamate iontophoresis. These small, slow spikes were abolished by blocking voltage-gated Ca2+ channels but not by blocking Na+ channels or N-methyl-D-aspartate receptors. We conclude that these Ca2+ spikes occurred in a spatially restricted region of the dendrite and were not actively propagated to the soma. In the presence of 10 mM tetraethylammonium chloride, the amplitudes of the iontophoretically evoked Ca2+ spikes were large, similar to those of the Ca2+ spikes evoked by somatic current injection, but their apparent thresholds were 63% lower. We conclude that dendritic K+ channels normally prevent the active propagation of Ca2+ spikes along the dendrite. In 36% of recorded cells dendritic glutamate iontophoresis evoked a Na+ spike with an apparent threshold 63% lower than those evoked by somatic current injection or somatic glutamate iontophoresis. Blockade of these low-threshold Na+ spikes by pharmacological or electrophysiological means often revealed underlying small dendritic Ca2+ spikes. When cells displaying the low-threshold Na+ spikes were voltage clamped at membrane potentials that prevented firing of the soma or axon, corresponding tetrodotoxin-sensitive current spikes could be evoked all-or-none by dendritic depolarization. We conclude that these low-threshold Na+ spikes were initiated in the dendrite, probably by local Ca2+ spikes, and subsequently propagated actively to the soma. Most cells displaying dendritic Na+ spikes fired multiple bursts of action potentials during tonic dendritic depolarization, whereas somatic depolarization of the same cells evoked only regular firing. We discuss the implications of dendritic Ca2+ and Na+ spikes for synaptic integration and neural input-output relations.

Action potential initiation and backpropagation in neurons of the mammalian CNS

Most neurons in the mammalian CNS encode and transmit information via action potentials. Knowledge of where these electrical events are initiated and how they propagate within neurons is therefore fundamental to an understanding of neuronal function. While work from the 1950s suggested that action potentials are initiated in the axon, many subsequent investigations have suggested that action potentials can also be initiated in the dendrites. Recently, experiments using simultaneous patch-pipette recordings from different locations on the same neuron have been used to address this issue directly. These studies show that the site of action potential initiation is in the axon, even when synaptic activation is powerful enough to elicit dendritic electrogenesis. Furthermore, these and other studies also show that following initiation, action potentials actively backpropagate into the dendrites of many neuronal types, providing a retrograde signal of neuronal output to the dendritic tree.

Axonal action-potential initiation and Na+ channel densities in the soma and axon initial segment of subicular pyramidal neurons

A long-standing hypothesis is that action potentials initiate first in the axon hillock/initial segment (AH-IS) region because of a locally high density of Na+ channels. We tested this idea in subicular pyramidal neurons by using patch-clamp recordings in hippocampal slices. Simultaneous recordings from the soma and IS confirmed that orthodromic action potentials initiated in the axon and then invaded the soma. However, blocking Na+ channels in the AH-IS with locally applied tetrodotoxin (TTX) did not raise the somatic threshold membrane potential for orthodromic spikes. TTX applied to the axon beyond the AH-IS (30-60 microm from the soma) raised the apparent somatic threshold by approximately 8 mV. We estimated the Na+ current density in the AH-IS and somatic membranes by using cell-attached patch-clamp recordings and found similar magnitudes (3-4 pA/microm2). Thus, the present results suggest that orthodromic action potentials initiate in the axon beyond the AH-IS and that the minimum threshold for spike initiation of the neuron is not determined by a high density of Na+ channels in the AH-IS region.

Modeling the attenuation and failure of action potentials in the dendrites of hippocampal neurons

We modeled two different mechanisms, a shunting conductance and a slow sodium inactivation, to test whether they could modulate the active propagation of a train of action potentials in a dendritic tree. Computer simulations, using a compartmental model of a pyramidal neuron, suggest that each of these two mechanisms could account for the activity-dependent attenuation and failure of the action potentials in the dendrites during the train. Each mechanism is shown to be in good qualitative agreement with experimental findings on somatic or dendritic stimulation and on the effects of hyperpolarization. The conditions under which branch point failures can be observed, and a few experimentally testable predictions, are presented and discussed.

Modeling back propagating action potential in weakly excitable dendrites of neocortical pyramidal cells

Simultaneous recordings from the soma and apical dendrite of layer V neocortical pyramidal cells of young rats show that, for any location of current input, an evoked action potential (AP) always starts at the axon and then propagates actively, but decrementally, backward into the dendrites. This back-propagating AP is supported by a low density (-gNa = approximately 4 mS/cm2) of rapidly inactivating voltage-dependent Na+ channels in the soma and the apical dendrite. Investigation of detailed, biophysically constrained, models of reconstructed pyramidal cells shows the following. (i) The initiation of the AP first in the axon cannot be explained solely by morphological considerations; the axon must be more excitable than the soma and dendrites. (ii) The minimal Na+ channel density in the axon that fully accounts for the experimental results is about 20-times that of the soma. If -gNa in the axon hillock and initial segment is the same as in the soma [as recently suggested by Colbert and Johnston [Colbert, C. M. & Johnston, D. (1995) Soc. Neurosci. Abstr. 21, 684.2]], then -gNa in the more distal axonal regions is required to be about 40-times that of the soma. (iii) A backward propagating AP in weakly excitable dendrites can be modulated in a graded manner by background synaptic activity. The functional role of weakly excitable dendrites and a more excitable axon for forward synaptic integration and for backward, global, communication between the axon and the dendrites is discussed.

Pattern and inhibition-dependent invasion of pyramidal cell dendrites by fast spikes in the hippocampus in vivo

The invasion of sodium spikes from the soma into dendrites was studied in hippocampal pyramidal cells by simultaneous extracellular and intracellular recordings in anesthetized rats and by simultaneous extracellular recordings of the somatic and dendritic potentials in freely behaving animals. During complex-spike patterns, recorded in the immobile or sleeping animal, dendritic invasion of successive spikes was substantially attenuated. Complex-spike bursts occurred in association with population discharge of CA3-CA1 pyramidal cells (sharp wave field events). Synaptic inhibition reduced the amplitude of sodium spikes in the dendrites and prevented the occurrence of calcium spikes. These findings indicate that (i) the voltage-dependent calcium influx into the dendrites is under the control of inhibitory neurons and (ii) the temporal coincidence of synaptic depolarization and activation of voltage-dependent calcium conductances by the backpropagating spikes during sharp wave bursts may be critical for synaptic plasticity in the intact hippocampus.

Influence of dendritic structure on firing pattern in model neocortical neurons

Neocortical neurons display a wide range of dendritic morphologies, ranging from compact arborizations to highly elaborate branching patterns. In vitro electrical recordings from these neurons have revealed a correspondingly diverse range of intrinsic firing patterns, including non-adapting, adapting and bursting types. This heterogeneity of electrical responsivity has generally been attributed to variability in the types and densities of ionic channels. We show here, using compartmental models of reconstructed cortical neurons, that an entire spectrum of firing patterns can be reproduced in a set of neurons that share a common distribution of ion channels and differ only in their dendritic geometry. The essential behaviour of the model depends on partial electrical coupling of fast active conductances localized to the soma and axon and slow active currents located throughout the dendrites, and can be reproduced in a two-compartment model. The results suggest a causal relationship for the observed correlations between dendritic structure and firing properties and emphasize the importance of active dendritic conductances in neuronal function.

Slow inactivation of Na+ current and slow cumulative spike adaptation in mouse and guinea-pig neocortical neurones in slices

1. Spike adaptation of neocortical pyramidal neurones was studied with sharp electrode recordings in slices of guinea-pig parietal cortex and whole-cell patch recordings of mouse somatosensory cortex. Repetitive intracellular stimulation with 1 s depolarizing pulses delivered at intervals of < 5 s caused slow, cumulative adaptation of spike firing, which was not associated with a change in resting conductance, and which persisted when Co2+ replaced Ca2+ in the bathing medium. 2. Development of slow cumulative adaptation was associated with a gradual decrease in maximal rates of rise of action potentials, a slowing in the post-spike depolarization towards threshold, and a positive shift in the threshold voltage for the next spike in the train; maximal spike repolarization rates and after-hyperpolarizations were unchanged. 3. The data suggested that slow adaptation reflects use-dependent removal of Na+ channels from the available pool by an inactivation process which is much slower than fast, Hodgkin-Huxley-type inactivation. 4. We therefore studied the properties of Na+ channels in layer II-III mouse neocortical cells using the cell-attached configuration of the patch-in-slice technique. These had a slope conductance of 18 +/- 1 pS and an extrapolated reversal potential of 127 +/- 6 mV above resting potential (Vr) (mean +/- S.E.M.; n = 5). Vr was estimated at -72 +/- 3 mV (n = 8), based on the voltage dependence of the steady-state inactivation (h infinity) curve. 5. Slow inactivation (SI) of Na+ channels had a mono-exponential onset with tau on between 0.86 and 2.33 s (n = 3). Steady-state SI was half-maximal at -43.8 mV and had a slope of 14.4 mV (e-fold)-1. Recovery from a 2 s conditioning pulse was bi-exponential and voltage dependent; the slow time constant ranged between 0.45 and 2.5 s at voltages between-128 and -68 mV. 6. The experimentally determined parameters of SI were adequate to simulate slow cumulative adaptation of spike firing in a single-compartment computer model. 7. Persistent Na+ current, which was recorded in whole-cell configuration during slow voltage ramps (35 mV s-1), also underwent pronounced SI, which was apparent when the ramp was preceded by a prolonged depolarizing pulse.