The electro-dynamics of the dendritic space in Purkinje cells of the cerebellum.

The functional geometry of the reconstructed dendritic arborization of Purkinje neurons is the object of this work. The combined effects of the local geometry of the dendritic branches and of the membrane mechanisms are computed in passive configuration to obtain the electrotonic structure of the arborization. Steady-currents applied to the soma and expressed as a function of the path distance from the soma form different clusters of profiles in which dendritic branches are similar in voltages and current transfer effectiveness. The locations of the different clusters are mapped on the dendrograms and 3D representations of the arborization. It reveals the presence of different spatial dendritic sectors clearly separated in 3D space that shape the arborization in ordered electrical domains, each with similar passive charge transfer efficiencies. Further simulations are performed in active configuration with a realistic cocktail of conductances to find out whether similar spatial domains found in the passive model also characterize the active dendritic arborization. During tonic activation of excitatory synaptic inputs homogeneously distributed over the whole arborization, the Purkinje cell generates regular oscillatory potentials. The temporal patterns of the electrical oscillations induce similar spatial sectors in the arborization as those observed in the passive electrotonic structure. By taking a video of the dendritic maps of the membrane potentials during a single oscillation, we demonstrate that the functional dendritic field of a Purkinje neuron displays dynamic changes which occur in the spatial distribution of membrane potentials in the course of the oscillation. We conclude that the branching pattern of the arborization explains such continuous reconfiguration and discuss its functional implications.

Spatial reconfiguration of charge transfer effectiveness in active bistable dendritic arborizations.

The aim of this work was to explore the electrical spatial profile of the dendritic arborization during membrane potential oscillations of a bistable motoneuron. Computational simulations provided the spatial counterparts of the temporal dynamics of bistability and allowed simultaneous depiction the electrical states of any sites in the arborization. We assumed that the dendritic membrane had homogeneously distributed specific electrical properties and was equipped with a cocktail of passive extrasynaptic and NMDA synaptic conductances. The electrical conditions for evoking bistability in a single isopotential compartment and in a whole dendritic arborization were computed and showed differences, revealing a crucial effect of dendritic geometry. Snapshots of the whole arborization during bistability revealed the spatial distribution of the density of the transmembrane current generated at the synapses and the effectiveness of the current transfer from any dendritic site to the soma. These functional maps changed dynamically according to the phase of the oscillatory cycle. In the low depolarization state, the current density was low in the proximal dendrites and higher in the distal parts of the arborization while the transfer effectiveness varied in a narrow range with small differences between proximal and distal dendritic segments. When the neuron switched to high depolarization state, the current density was high in the proximal dendrites and low in the distal branches while a large domain of the dendritic field became electrically disconnected beyond 200 micro m from the soma with a null transfer efficiency. These spatial reconfigurations affected dynamically the size and shape of the functional dendritic field and were strongly geometry-dependent.

Dendritic spatial flicker of local membrane potential due to channel noise and probabilistic firing of hippocampal neurons in culture.

Whole-cell recordings and imaging of dissociated hippocampal neurons stained with voltage sensitive dye provide a new microscopic picture of neuronal excitation. This is the first attempt to combine imaging of active channel clusters on the geometry of live neurons and a theoretical approach. During single somatic action potentials and the back-invasion into the neurites, local mean potentials are generated at sites of active channel clusters which are unevenly distributed in the neuronal membrane. Similar mean membrane potentials are observed in the neurites and at the soma. Identical action potentials produce different spatial patterns of mean membrane potentials from trial to trial. This spatial variability is explained by the stochastic behavior of the channels in the clusters. When hippocampal neurons are excited by synaptic inputs, their evoked responses are probabilistic and generate variable spatial patterns of mean membrane potential trial after trial. Our stochastic model reproduces this random behavior by assuming that the voltage fluctuations generated by channel noise are added to the synaptic potentials reaching the soma. We demonstrate that the probability of action potential initiation depends on the strength of the synaptic input, the diameter of the dendrites and the relative positions of the channel clusters, of the synapse and of the soma.

Imaging stochastic spatial variability of active channel clusters during excitation of single neurons.

Topographical maps of membrane voltages were obtained during action potentials by imaging, at 1 microm resolution, live dissociated neurons stained with the voltage sensitive dye RH237. We demonstrate with a theoretical approach that the spatial patterns in the images result from the distribution of net positive charges condensed in the inner sites of the membrane where clusters of open ionic channels are located. We observed that, in our biological images, this spatial distribution of open channels varies randomly from trial to trial while the action potentials recorded by the microelectrode display similar amplitudes and time-courses. The random differences in size and intensity of the spatial patterns in the images are best evidenced when the time of observation coincides with the duration of single action potentials. This spatial variability is explained by the fact that only part of the channel population generates an action potential and that different channels open in turn in different trials due to their stochastic operation. Such spatial flicker modifies the direction of lateral current along the neuronal membrane and may have important consequences on the intrinsic processing capabilities of the neuron.

Activity-dependent reconfiguration of the effective dendritic field of motoneurons.

A neuron in vivo receives a continuous bombardment of synaptic inputs that modify the integrative properties of dendritic arborizations by changing the specific membrane resistance (R(m)). To address the mechanisms by which the synaptic background activity transforms the charge transfer effectiveness (T(x)) of a dendritic arborization, the authors simulated a neuron at rest and a highly excited neuron. After in vivo identification of the motoneurons recorded and stained intracellularly, the motoneuron arborizations were reconstructed at high spatial resolution. The neuronal model was constrained by the geometric data describing the numerized arborization. The electrotonic structure and T(x) were computed under different R(m) values to mimic a highly excited neuron (1 kOhm x cm(2)) and a neuron at rest (100 kOhm x cm(2)). The authors found that the shape and the size of the effective dendritic fields varied in the function of R(m). In the highly excited neuron, the effective dendritic field was reduced spatially by switching off most of the distal dendritic branches, which were disconnected functionally from the somata. At rest, the entire dendritic field was highly efficient in transferring current to the somata, but there was a lack of spatial discrimination. Because the large motoneurons are more sensitive to variations in the upper range of R(m), they switch off their distal dendrites before the small motoneurons. Thus, the same anatomic structure that shrinks or expands according to the background synaptic activity can select the types of its synaptic inputs. The results of this study demonstrate that these reconfigurations of the effective dendritic field of the motoneurons are activity-dependent and geometry-dependent.

The problem of the morphological noise in reconstructed dendritic arborizations.

For technical, instrumental and operator-related reasons, three-dimensional (3D) reconstructions of neurons obtained from intracellularly stained neuronal pieces scattered in serial sections are blurred by some morphological noise. This noise may strongly invalidate conclusions drawn from models built using the 3D reconstructions and it must be taken into account when retrieving digitized neurons from available databases. We analyse on several vertebrate neurons examples the main noise-generating sources and the consequences of the noise on the 'quality' of the data. We show how the noise can be detected and evaluated in any database, if sufficient information is presented in this database.

Differential back-invasion of a single complex dendrite of an abducens motoneuron by N-methyl-D-aspartate-induced oscillations: a simulation study.

Intracellular recording of abducens motoneurons in vivo has shown that ionophoretic applications of N-methyl-D-aspartate produced long-lasting membrane potential oscillations including a slow depolarization plateau with a burst of fast action potentials. This complex N-methyl-D-aspartate pattern was reproduced in the model of abducens motoneuron in vivo identified, intracellularly stained with horseradish peroxidase and reconstructed at high spatial resolution. The excitable soma of the simulated cell contained voltage-gated Ca, Na and K conductances, N-methyl-D-aspartate-gated voltage-sensitive Ca-Na-K conductance and Ca-dependent K conductance. The dendrite was passive either completely or with the exception of branching nodes containing N-methyl-D-aspartate conductances of the same slow kinetics but of lower values than at the soma. In the completely passive case, the N-methyl-D-aspartate pattern decayed with different rates along different dendritic paths depending on the geometry and topology of the reconstructed dendrite. The branches formed four clusters discriminated in somatofugal attenuations of steady voltages, and were correspondingly discriminated in attenuation of the complex N-methyl-D-aspartate pattern. Fast spikes decayed more than the slow depolarization plateau so that the prevalence of slow over fast components in the transformed pattern increased with somatofugal path distance. As a consequence, the lower the electrotonic effectiveness of a branch in the cluster or in the whole arborization, the lower both the voltage level and the frequency range of its voltage modulation by N-methyl-D-aspartate oscillations. In the case of active branching points, the somatic pattern changed depending on the level of activation of dendritic N-methyl-D-aspartate conductances with slow kinetics of voltage sensitivity. The higher this level, the longer the plateau and burst, and the greater the discharge rate; and the spikes in the burst were smaller. When the pattern spread in the dendrite, the fast spikes decayed and the slow plateau was boosted, with a greater effect along the somatofugal path containing more branching points. These results show how the somatofugal back-invasion along the dendrites by activity patterns generated at the soma can tune voltage-sensitive dendritic conductances. The dendritic back-invasion is geometry- and topology-dependent. It is proposed as a subtle feedback mechanism for the neuron to control its own synaptic inputs.

Local mechanisms of phase-dependent postsynaptic modifications of NMDA-induced oscillations in the abducens motoneurons: a simulation study.

1. In vivo experiments have shown that extracellular microelectrophoretic application of N-methyl-D-aspartate (NMDA) induced oscillatory plateau potentials with bursts of action potentials in rat abducens motoneurons. The period of these slow NMDA oscillations could be altered by single trigeminal non-NMDA excitatory input delivered at low frequency during the NMDA oscillations. 2. A resetting of the oscillations was observed depending on the phase of slow oscillatory cycle during which the trigeminal excitation occurred. 3. We investigated local mechanisms responsible for the phase-dependent modifications of NMDA oscillations, including contributions of voltage and concentration transients, in the mathematical model of the isopotential membrane compartment equipped with voltage-gated Na+, K+, and Ca2+ channels, with Ca2+-dependent K+ channels, and with ligand-gated NMDA and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor channels. The faithful model was constructed with the use of models described earlier, which were modified by increasing time constants of kinetic variables of all voltage-gated conductances and by including coupled dynamics of voltages and ion concentrations. The changes in ion concentrations were produced near the membrane by transmembrane currents and removal mechanisms (pumps, diffusion). 4. This work focuses on local arrangement of voltage- and ligand-gated conductances and on local ion concentration changes in two separate pools: the postsynaptic pool of AMPA receptors and the extrasynaptic pool. In terms of the electrotonic and diffusional length constants, these pools were electrotonically close but diffusionally remote. 5. It was found that the effect of resetting can be produced by a local interaction between plateau and spike-generating conductances and glutamate receptors. 6. In vivo phase-dependent interactions between NMDA oscillations and AMPA synaptic input were reproduced by the local model only when changes in intracellular sodium and extracellular potassium concentrations were taken into account and the mechanisms of ion removal from postsynaptic pools had slower kinetics than the fast pump system operating in the extracellular pool. 7. Postsynaptic changes in ion concentrations of Na+ and K+ in intra- and extracellular layers near the membrane shift of Nernst equilibrium potentials for these ions depending on the phase of activation of synaptic input. Thus Na+ and k+ components of all transmembrane currents involved in the pattern generation are differently affected by synaptic action during the oscillations. We conclude that slow postsynaptic changes in ion concentrations near the membrane play a key role in the resetting of the NMDA oscillations.

Fluorescence imaging of local membrane electric fields during the excitation of single neurons in culture.

The spatial distribution of depolarized patches of membrane during the excitation of single neurons in culture has been recorded with a high spatial resolution (1 micron2/pixel) imaging system based on a liquid-nitrogen-cooled astronomical camera mounted on an inverted microscope. Images were captured from rat nodose neurons stained with the voltage-sensitive dye RH237. Conventional intracellular microelectrode recordings were made in synchrony with the images. During an action potential the fluorescence changes occurred in localized, unevenly distributed membrane areas, which formed clusters of depolarized sites of different sizes and intensities. When fast conductances were blocked by the addition of tetrodotoxin, a reduction in the number and the intensities of the depolarized sites was observed. The blockade by tetrodotoxin of voltage-clamped neurons also reduced the number of depolarized sites, although the same depolarizing voltage step was applied. Similarly, when a voltage-clamped neuron was depolarized by a constant-amplitude voltage step, the number of depolarized sites varied according to the degree of activation of the voltage-sensitive channels, which was modified by changing the holding potential. These results suggest that the spatial patterns of depolarization observed during excitation are related to the operations of ionic channels in the membrane.

Electrotonic clusters in the dendritic arborization of abducens motoneurons of the rat.

Following reconstruction with high spatial resolution of the 3-D geometry of the dendritic arborizations of two abducens motoneurons, we simulated the distribution of electronic voltage over the whole dendritic tree. Here, we demonstrate that the complex stochastic electronic structure of both motoneurons can be reduced to a statistically significant small set of well discriminated clusters. These clusters are formed by dendritic branches belonging to different dendrites of the neuron but with similar electronic properties. A cluster analysis was performed to estimate quantitatively the partition of the branches between the dendritic clusters. The contents of the clusters were analysed in relation to their stability under different values of specific membrane resistivity (Rm), to their remoteness from the soma and their location in 3-D space. The cluster analysis was executed in a 2-D parameter space in which each dendritic branch was described by the mean electrotonic voltage and gradient. The number of clusters was found to be four for each motoneuron when computations were made with Rm = 3 k omega.cm2. An analysis of the cluster composition under different Rm revealed that each cluster contained invariant and variant branches. Mapping the clusters upon the dendritic geometry of the arborizations allowed us to describe the cluster distribution in terms of the 3-D space domain, the 2-D path distance domain and the total surface area of the tree. As the cluster behaviour reflects both the geometry and the changes in the neuronal electrotonic structure, we conclude that cluster analysis provides a tool to handle the functional complexity of the arborizations without losing relevant information. In terms of synaptic activities, the stable dendritic branches in each cluster may process the synaptic inputs in a similar manner. The high percentage of stable branches indicates that geometry is a major factor of stability for the electrotonic clusters. Conversely, the variant branches introduce the conditions for mechanisms of functional postsynaptic plasticity.

Stochastic geometry and electronic architecture of dendritic arborization of brain stem motoneuron.

We describe how the stochastic geometry of dendritic arborization of a single identified motoneuron of the rat affects the local details of its electrotonic structure. After describing the 3D dendritic geometry at high spatial resolution, we simulate the distribution of voltage gradients along dendritic branches under steady-state and transient conditions. We show that local variations in diameters along branches and asymmetric branchings determine the non-monotonous features of the heterogeneous electrotonic structure. This is defined by the voltage decay expressed as a function of the somatofugal paths in physical distances (voltage gradient). The fan-shaped electrotonic structure demonstrates differences between branches which are preserved when simulations are computed from different values of specific membrane resistivity although the absolute value of their voltages is changed. At given distances from soma and over long paths, some branches display similar voltages resulting in their grouping which is also preserved when specific membrane resistivity is changed. However, the mutual relation between branches inside the group is respecified when different values of specific membrane resistivity are used in the simulations. We find that there are some invariant features of the electrotonic structure which are related to the geometry and not to the electrical parameters, while other features are changed by altering the electrical parameters. Under transient conditions, the somatofugal invasion of the dendritic tree by a somatic action potential shifts membrane potentials (above 10 mV) of dendritic paths for unequal distances from the soma during several milliseconds. Electrotonic reconfigurations and membrane shifts might be a mechanism for postsynaptic plasticity.

[Quantitative imaging of the heterogeneity of membrane activation of mammalian neurons and glial cells]. / Visualisation quantitative de l'hétérogénéité de l'activation membranaire des neurones et de la glie de mammifère.

By using quantitative imaging with an ultra-high sensitivity, it was possible to observe the simultaneous action of multiple patches unevenly distributed over the membranes of neurons and glial cells in culture. We used a voltage-sensitive probe to stain vitally the cells. The instrumentation consisted of a liquid-nitrogen cooled matrix of 222,530 photodetectors with a spatial resolution of 0.25 microns 2, a photodynamic range of 10(5), a detection level of a few tens of photons and a maximum time resolution of 500 microseconds. Electrical and pharmacological stimulations were applied to produce the activation of the cells which was accompanied by large variations of the level of fluorescence, giving a precise spatial localization of active domains over the soma-neuritic membranes. These images of fluorescent signals are interpreted as corresponding to the plasmalemmal localization of voltage-dependent channels. This finding, which had not been previously observed with voltage-sensitive probes in fluorescent dye imaging indicates the possibility of measuring the activity of independently functioning domains in single neurons.

The dendrites of single brain-stem motoneurons intracellularly labelled with horseradish peroxidase in the cat. Morphological and electrical differences.

The geometrical differences between individual dendrites of a given motoneuron were investigated in the cat. We chose two brain-stem motoneurons involved in different motor activities. One abducens and one laryngeal motoneuron were selected from two series of experiments which had combined intracellular recording and horseradish peroxidase staining. Three-dimensional reconstructions were made using a computer-aided microscope to obtain high-resolution measurements from serial histological sections. Each dendrite was characterized by computer dissection. Comparisons between dendrites were made on the basis of the following parameters: spatial projections, length, diameters, tapering, branching pattern, daughter--branch ratio and branching power. The present findings show that each dendrite projects to specific terminal fields for both motoneurons and are different in the complexity of their geometry and branching structure. The consequences of this complexity for the cable properties of the motoneurons were analysed. The dendrites of the two motoneurons were partitioned into a series of contiguous regions deemed short enough to be considered an isopotential cylinder and the steady-state properties were calculated for each segment. The properties of each segment were then combined for each dendrite for the following parameters: electronic distance, somatopetal and somatofugal voltage attenuation, input resistance and charge transfer effectiveness ratio. The present results show significant differences in the electrical behaviour of individual dendrites. Branch-to-branch computation reveals low attenuation pathways between branches suggesting the possibility of local influences within the distal branches of the dendritic arborization. It is proposed that the individual dendrites of the motoneuron function as distinct channels and/or integrators for afferent inputs.

The dendrites of single brain-stem motoneurons intracellularly labelled with horseradish peroxidase in the cat. An ultrastructural analysis of the synaptic covering and the microenvironment.

Two laryngeal motoneurons intracellularly stained with horseradish peroxidase were studied ultrastructurally. The precise position of the ultrastructural observations made along the dendrites was obtained from the computer-reconstruction of the motoneurons in three dimensions. The shape and the size of the synaptic boutons, the percentage of membrane covered by bouton appositions and active zones, the number of boutons per 100 microns2 (packing density) were analysed on the soma and on the labelled dendrites at different distances from the soma up to 1000 microns. The results revealed no important regional differences in the mean length of synaptic apposition. The packing density was in the range of 9.3-14.9 boutons per 100 microns2 and was not correlated with the distance from the soma. The percentage apposition covering was higher on the soma and the proximal part of the dendrites than on the remaining part of the dendritic arborization. Close appositions between labelled dendrite and unlabelled somata and/or dendrites together with dendro-dendritic synapses suggested the possibility that the dendrites may be involved in local cell-to-cell communication. Microdendrites emerging from the soma or the proximal dendrites were contacted by synaptic boutons which may be more efficient as revealed by computation.

L-glutamate and N-methyl-D-asparatate actions on membrane potential and conductance of cat abducens motoneurones.

The actions of L-glutamate and N-methyl-D-aspartate (NMDA) were studied with intracellular recordings from cat abducens motoneurones. Amino acids were electrophoresed extracellularly from the same 7-barreled electrode types as those used in the spinal cord. Depolarization development, conductance changes and firing pattern induced by amino acids were very similar to those described for spinal motoneurones. The shape of NMDA depolarization suggests a uniform distribution of the involved receptors on the membrane of the motoneurone.

Histochemical identification of two fibre types in the retractor bulbi muscle of the cat.

In adult cats, the fibre population of the retractor bulbi muscle (RB) was analysed, using the histochemical reactions of ATPases. The muscle was found to contain type-2 fibres only, of which 70% were 2a and 30% 2b. Such ATPase profiles, corresponding to fast-twitch fibers, are in agreement with the mechanical properties of the muscle. Both types, 2a and 2b, included fibres in which the oxidative enzyme content was high and fibres in which it was low. The glycogen content of all fibers in the RB was uniformly low.

Resistance to glycogen depletion of motor units in the cat rectus lateralis muscle.

In nembutal anesthetized adult cats, intracellular stimulation of single abducens motoneurones was used to elicit glycogen depletion of their muscle units. Stimulation by short trains (13 pulses at 40 Hz) delivered once a second, was applied for 20 to 110 min. The activation of the motor unit was monitored by intracellular recording of motoneurone action potentials and by EMG. After the end of stimulation, the muscle was excised and frozen to be cut in serial sections that were processed for demonstration of either glycogen, ATPases or SDH. In two experiments, a motor unit could be histochemically identified because 10-15 fibres showed zones of complete glycogen depletion measuring about 5 mm in length. All the depleted fibres had the same histochemical profile: ATPases reactions gave dark staining with alkaline preincubation and light staining with acid preincubation whereas SDH activity was low. In other experiments, prolonged stimulation produced either no depletion at all or very limited zones of partial depletion in a few muscle fibres.

On re-excitation of feline motoneurones: its mechanism and consequences.

Conditions required for re-excitation of lumbosacral motoneurones, i.e. for double impulses in the motor axons associated with a single soma-dendritic action potential, were examined in cats anaesthetized with pentobarbitone and paralysed with gallamine triethiodide. Simultaneous recording from a motoneurone (intracellular, and in some experiments also extracellular), and from its axon in a ventral root, was used to assess the relations between the soma and the double axonal action potentials. Action potentials (greater than 70 mV) evoked by brief depolarizing current pulses applied intracellularly were never observed to cause re-excitation. Re-excitation could, however, be regularly induced by procedures which increased the delay between the initial segment and soma-dendritic components of these potentials. Re-excitation could be evoked (i) when brief hyperpolarizing pulses were applied before the onset of the soma-dendritic spikes, (ii) when the depolarizing pulses were applied on a background of long hyperpolarizing pulses or (iii) when two action potentials were evoked in a quick succession (by two brief depolarizing pulses). No relationship was found between the presence of re-excitation of motor axons and the presence of the delayed depolarization which follows the soma-dendritic spikes. Neither re-excitation nor delayed depolarization were found to be dependent upon re-excitation of the initial segment. These observations are thus at variance with previous suggestions that the initial segment spikes induce the re-excitation of motor axons and that the initial segment spikes cause the delayed depolarization following soma-dendritic spikes. Since re-excitation of a motor axon occurred without any signs of a second initial segment spike, it is concluded that it is initiated at the level of the axon, most likely at the first node of Ranvier. Re-excitation of motor axons was also observed during repetitive firing induced by intracellular current injection. However, it occurred then only occasionally, and only under strong depolarizing drive. It is thus not expected to be a common phenomenon under natural conditions of repetitive firing.

Ultrastructural and electrophysiological properties of accessory abducens nucleus motoneurones: an intracellular horseradish peroxidase study in the cat.

The physiological and morphological (light and electron microscopy) properties of six retractor bulbi motoneurones were analysed using the technique of intracellular recording and intracellular labelling with horseradish peroxidase. The retractor bulbi motoneurones were identified by antidromic invasion and orthodromic responses following stimulation of trigeminal afferents were studied. Two of these motoneurones were examined ultrastructurally. Terminal boutons forming synapses with labelled soma, labelled proximal and distal dendrites were characterized. Serial sections allowed the axon hillock to be analyzed and the initial segment of a presumed motoneurone to be observed in the section where the injected motoneurone was described. The ultrastructure of unidentified elements observed in the accessory abducens nucleus is stressed.

Comparison of antidromic and orthodromic action potentials of identified motor axons in the cat's brain stem.

Recordings were made from identified central axons at a known distance from their somata, to compare the action potentials resulting from antidromic and synaptic excitation. By taking advantage of the anatomical configuration within the brain stem of the motoneurones innervating the retractor bulbi muscle in the orbit, their axons were penetrated in the VIth nucleus and labelled by electrophoretic injection of horseradish peroxidase. Excitatory post-synaptic potentials recorded in the retractor bulbi axons at about 3 mm from the soma were six times smaller than in the soma. The space constant of the axonal segment between the retractor bulbi and the abducens nucleus was estimated to be 1.7 mm. When the axons propagated action potentials the attenuation was increased to eighteen times due to the nodes of Ranvier intercalated between the soma and the site of recording. Antidromic action potentials displayed stepwise changes in amplitude and shape when stimuli were applied at intervals decreasing from 5 ms to 0.7 ms. The changes were related to the different lengths of refractoriness of the soma, initial segment and axon. Orthodromic action potentials evoked by synaptic excitation displayed similar changes in amplitude and shape. These observations lead to the conclusion that the soma, initial segment and neighbouring nodes of Ranvier contribute significantly to the shape of the action potential. Contrary to the generally accepted view, it appears that the efferent discharge along motor axons can be initiated without a simultaneous activation of the somato-dendritic or even the initial segment membrane, as revealed by the lack of somato-dendritic and/or initial segment contribution to the shape of the synaptically evoked action potentials.