Presynaptic inhibition and antidromic discharges in crayfish primary afferents.

The mechanisms of presynaptic inhibition have been studied in sensory afferents of a stretch receptor in an in vitro preparation of the crayfish. Axon terminals of these sensory afferents display primary afferent depolarisations (PADs) mediated by the activation of GABA receptors that open chloride channels. Intracellular labeling of sensory axons by Lucifer yellow combined with GABA immunohistochemistry revealed the presence of close appositions between GABA-immunoreactive boutons and sensory axons close to their first branching point within the ganglion. Electrophysiological studies showed that GABA inputs mediating PADs appear to occur around the first axonal branching point, which corresponds to the area of transition between active and passive propagation of spikes. Moreover, this study demonstrated that whilst shunting appeared to be the sole mechanism involved during small amplitude PADs, sodium channel inactivation occurred with larger amplitude PADs. However, when the largest PADs (>25 mV) are produced, the threshold for spike generation is reached and antidromic action potentials are elicited. The mechanisms involved in the initiation of antidromic discharges were analyzed by combining electrophysiological and simulation studies. Three mechanisms act together to ensure that PAD-mediated spikes are not conveyed distally: 1) the lack of active propagation in distal regions of the sensory axons; 2) the inactivation of the sodium channels around the site where PADs are produced; and 3) a massive shunting through the opening of chloride channels associated with the activation of GABA receptors. The centrally generated spikes are, however, conveyed antidromically in the sensory nerve up to the proprioceptive organ, where they inhibit the activity of the sensory neurons for several hundreds of milliseconds.     

Presynaptic action potential amplification by voltage-gated Na+ channels in hippocampal mossy fiber boutons

Action potentials in central neurons are initiated near the axon initial segment, propagate into the axon, and finally invade the presynaptic terminals, where they trigger transmitter release. Voltage-gated Na(+) channels are key determinants of excitability, but Na(+) channel density and properties in axons and presynaptic terminals of cortical neurons have not been examined yet. In hippocampal mossy fiber boutons, which emerge from parent axons en passant, Na(+) channels are very abundant, with an estimated number of approximately 2000 channels per bouton. Presynaptic Na(+) channels show faster inactivation kinetics than somatic channels, suggesting differences between subcellular compartments of the same cell. Computational analysis of action potential propagation in axon-multibouton structures reveals that Na(+) channels in boutons preferentially amplify the presynaptic action potential and enhance Ca(2+) inflow, whereas Na(+) channels in axons control the reliability and speed of propagation. Thus, presynaptic and axonal Na(+) channels contribute differentially to mossy fiber synaptic transmission.     

The generation of action potentials in the terminals of the Schaffer collaterals during long-term potentiation in the CA1 region of the hippocampus]

Experiments were performed in rat hippocampal slices. Activity of individual CA3 pyramidal neurons and field potentials in the CA1 areas were recorded extracellularly. The collision technique was applied to detect the antidromic origin of the background action potentials in the somata of CA3 neurons. Threshold stimulation of terminals of the Schaffer collaterals in the stratum radiatum of the CA1 area was applied to study their excitability during the CA1 long-term potentiation. During the long-term potentiation, antidromic action potentials appeared in the somata of the CA3 neurons. The obtained evidence suggests that the synaptic potentiation is accompanied by an enhancement of axon terminal excitability resulting in generation of the action potentials.     

Non-junctional modulation of neurogenic twitches of the guinea-pig ileum by some peptides and other compounds in the triple bath

The effects of some neuropeptide transmitter candidates and of some other neurotoxins or drugs on conduction of neural excitation were studied in myenteric plexus-longitudinal muscle strips from the guinea-pig ileum. A preparation in a special triple bath was drawn through two rubber membranes dividing the strip into three segments. Neurogenic stimulation of the oral segment set up nerve action potentials propagating aborally across the middle segment so that the aboral segment might also be invaded. Drugs were added to the middle segment to affect neuronal propagation (non-junctional effects) which was monitored by twitch amplitude of the aboral segment. The application of bradykinin and cromakalim did not affect aboral twitches although strong contractile and relaxatory effects were observed when the drugs were applied directly to the aboral segment; no neurogenic effects thus manifested. Capsaicin and neurotensin, when applied both to the middle and aboral segments, elevated the tone of the preparations accompanied with a decrease in twitch amplitude; these effects may have been due to neurogenic stimulation and release of other motor neurotransmitters. The application of VIP, apamin and dendrotoxin to the middle as well as to the aboral segments augmented aboral twitches, which might be at least partly due to facilitation of nerve action potential propagation in nerve terminals of cholinergic motor fibres.     

Squid giant axons. A model for the neuron soma?

Insertion of electrically floating wires along the axis of a squid giant axon produces an apparent increase in diameter in the region where the wire surface has been treated to give it a low resistance. The shape of action potentials propagating into this region depend upon the surface resistance (and the length) of the wire. As this segment's internal resistance is lowered by reducing the wire's surface resistance, the following characteristic sequence of changes in the action potential is seen at the transition region: (a) the duration increases; (b) two peaks develop, the first one generated in the normal axon region and the second one generated later in the axial wire region, and; (c) blockage occurs (for a very low resistance wire). Action potentials recorded at the membrane region near the tip of the axial wire in (b) resemble those recorded at the initial segment of neurons upon antidromic invasions. Squid axon action potentials propagated from a normal region into that containing the low resistance wire also resemble antidromic invasions recorded in neuron somas. Hyperpolarizing current pulses applied through the wire act as if the wire surface resistance was momentarily reduced. For example, the two components of the action potential recorded at the axial wire membrane region noted in (b) can be sequentially blocked by the application of increasing hyperpolarizing current through the wire. Similar effects are seen when hyperpolarizing currents are injected into motoneuron somas. It is concluded that the geometrical properties of the junction of a neuron axon with its soma may be in themselves sufficient to determine the shape of the action potentials usually recorded by microelectrodes.     

Antidromic discharges of dorsal root afferents in the neonatal rat.

Presynaptic inhibition of primary afferents can be evoked from at least three sources in the adult animal: 1) by stimulation of several supraspinal structures; 2) by spinal reflex action from sensory inputs; or 3) by the activity of spinal locomotor networks. The depolarisation in the intraspinal afferent terminals which is due, at least partly, to the activation of GABA(A) receptors may be large enough to reach firing threshold and evoke action potentials that are antidromically conducted into peripheral nerves. Little is known about the development of presynaptic inhibition and its supraspinal control during ontogeny. This article, reviewing recent experiments performed on the in vitro brainstem/spinal cord preparation of the neonatal rat, demonstrates that a similar organisation is present, to some extent, in the new-born rat. A spontaneous activity consisting of antidromic discharges can be recorded from lumbar dorsal roots. The discharges are generated by the underlying afferent terminal depolarizations reaching firing threshold. The number of antidromic action potentials increases significantly in saline solution with chloride concentration reduced to 50% of control. Bath application of the GABA(A) receptor antagonist, bicuculline (5-10 microM) blocks the antidromic discharges almost completely. Dorsal root discharges are therefore triggered by chloride-dependent GABA(A) receptor-mediated mechanisms; 1) activation of descending pathways by stimulation delivered to the ventral funiculus (VF) of the spinal cord at the C1 level; 2) activation of sensory inputs by stimulation of a neighbouring dorsal root; or 3) pharmacological activation of the central pattern generators for locomotion evokes antidromic discharges in dorsal roots. VF stimulation also inhibited the response to dorsal root stimulation. The time course of this inhibition overlapped with that of the dorsal root discharge suggesting that part of the inhibition of the monosynaptic reflex may be exerted at a presynaptic level. The existence of GABA(A) receptor-independent mechanisms and the roles of the antidromic discharges in the neonatal rat are discussed.     

Superposition of impulse activity in a rapidly-adapting afferent unit model

Mechanosensory afferent units consist of a parent axon, the peripheral axonal arborization, and the branch terminal mechanoreceptors. The present work uses a mathematical model to describe the contribution of a given number of rapidly-adapting mechanoreceptors to the impulse pattern of their parent axon. In the model impulses initiated by any driven mechanoreceptor instantaneously propagate orthodromically and antidromically. The model also incorporates the axonal absolute refractory period as well as ortho-and antidromically elicited recovery cycles. In separate computations, periodic or random (Poisson process) trains of short-duration stimuli at constant amplitude are delivered to a given number (N=2–30) of co-innervated mechanoreceptors. The superposition of component impulse trains always departs from the theoretical ideal (Poisson process). Such departures are attributable to: (i) the number of driven mechanoreceptors, when N is small, (ii) axonal absolute refractory period, during maximal amplitude stimulation, and (iii) antidromic recovery cycles as well as absolute refractoriness, during submaximal-amplitude stimulation. Computations reveal that this instantaneous reset model results in the elimination of information extracted by driven mechanoreceptors. Model predictions with Poisson stimulation at varied amplitudes are compared to G-hair afferent unit responses to analogous stimulation. Qualitatively opposite results with respect to parent axonal impulse patterns imply that the axonal arborization is not simply a substrate for impulse propagation from branch terminals to parent axon.     

Effect of a volatile anesthetic upon presynaptic excitability in mammalian hippocampus

Changes in presynaptic terminal axon excitability produced by enflurane in the rat hippocampal slice preparation were investigated by stimulation of Schaeffer collateral terminal axons and by recording single unit antidromic action potentials. Stimulating pulses were preceded by conditioning hyperpolarizing or depolarizing current pulses. A plot of net threshold for action potential generation against the conditioning pulse yields an "accommodation curve;" changes in this curve can be used to assess the mechanism by which changes in excitability are produced. Enflurane, at a concentration equivalent to approximately equal to 1.3 times the minimum alveolar concentration, reduced excitability of terminal axons and increased accommodation in a manner consistent with a possible change in the inactivation of gNa.     

Propagation of action potentials in squid giant axons. Repetitive firing at regions of membrane inhomogeneities

Effects of reduction in potassium conductance on impulse conduction were studied in squid giant axons. Internal perfusion of axons with tetraethylammonium (TEA) ions reduces G K and causes the duration of action potential to be increased up to 300 ms. This prolongation of action potentials does not change their conduction velocity. The shape of these propagating action potentials is similar to membrane action potentials in TEA. Axons with regions of differing membrane potassium conductances are obtained by perfusing the axon trunk and one of its two main branches with TEA after the second branch has been filled with normal perfusing solution. Although the latter is initially free of TEA, this ion diffuses in slowly. Up until a large amount of TEA has diffused into the second branch, action potentials in the two branches have very different durations. During this period, membrane regions with prolonged action potentials are a source of depolarizing current for the other, and repetitive activity may be initiated at transitional regions. After a single stimulus in either axon region, interactions between action potentials of different durations usually led to rebound, or a short burst, of action potentials. Complex interactions between two axon regions whose action potentials have different durations resembles electric activity recorded during some cardiac arrhythmias.     

Synaptic processes in the neurons of Clarke's column evoked by an antidromic volley from the dorsolateral column

In cats under nembutal-chloralose anaesthesia we investigated the response of neurons of Clarke's column to stimulation of axons ascending in the dorsal part of the lateral funiculus. Excitation of the descending fibers of the funiculus was prevented either by an ipsilateral hemisection of the thoracic cord carried out 7–10 days previously, which caused them to degenerate, or by stimulation of ascending axons in the region of the restiform bodies. It was found that with both kinds of stimulation records could be obtained from neurons in Clarke's column in which a descending volley causes not antidromic action potentials but primary excitatory postsynaptic potentials (EPSP). The length of the latent period of the EPSP (10–15 msec) suggests that they are monosynaptic. Such neurons may also be activated by low- or high-threshold afferents from various muscles; evidently they correspond to those described by Retheyi [14] as "edge" neurons on which terminate collaterals of axons ascending in the dorsal spinocerebellar tract (DSCT). In some of the neurons of the DSCT whose axons are distinguished by a low conduction velocity, stimulation of the dorsolateral funiculus caused not only antidromic spikes but also EPSP's following after them, and it would seem that the "edge" neurons were involved in their formation. We consider the possible functional role of a negative feed back loop formed by axon collaterals of neurons of the DSCT and by the "edge" neurons.A. A. Bogomolets Institute of Physiology, Academy of Sciences of the Ukrainian SSR, Kiev. Translated from Neirofiziologiya, Vol. 2, No. 3, pp. 269–278, May–June, 1970.     

中枢神经系统内神经元信号处理(英文)

一种新颖的轴突断端(axon bleb)膜片钳记录方法大力促进了中枢神经系统轴突功能的研究。我们的工作应用这一方法揭示了大脑皮层锥体神经元的数码信号(具全或无特性的动作电位)的爆发和传播机制。在轴突始段(axon initial segment,AIS)远端高密度聚集的低阈值Na+通道亚型Nav1.6决定动作电位的爆发;而在AIS近端高密度聚集的高阈值Na+通道亚型Nav1.2促进动作电位向胞体和树突的反向传播。应用胞体和轴突的同时记录,我们发现胞体阈下膜电位的变化可以在轴突上传播较长的距离并可到达那些离胞体较近的突触前终末。进一步的研究证明了胞体膜电位的变化调控动作电位触发的突触传递,该膜电位依赖的突触传递是一种模拟式的信号传递。轴突上一类特殊K+通道(Kv1)的活动调制动作电位的波形,特别是其波宽,从而调控各种突触前膜电位水平下突触强度的变化。突触前终末的背景Ca2+浓度也可能参与模拟信号的传递。这些发现深化了我们对中枢神经系统内神经信号处理基本原理的认识,进而帮助我们理解脑如何工作。     

Electrophysiological Study of Spatial Distribution of Vestibulospinal Neurons in the Frog Rana ridibunda

In experiments on a perfused brain preparation of the frog Rana ridibunda, the vestibulospinal neurons were identified, based on the excitatory postsynaptic potentials (EPSP) that appeared in response to an ipsilateral stimulation of the vestibular nerve and on the antidromic activity in response to stimulation of the cervical and lumbar enlargements of the spinal cord. The cells that could be antidromically activated only by stimulation of the cervical cord were designated as C-neurons. The cells that could be antidromically activated by stimulation of the lumbar cord were designated as L-neurons. The intracellular activity was recorded in 244 neurons of the vestibular nuclear complexes, out of which 127 cells (52%) were C-neurons and 117 (48%), L-neurons. The antidromic action potentials were recorded from the cells of lateral (143 neurons, 58.6%), descending (75 neurons, 30.7%), and medial (26 neurons, 10.6%) vestibular nuclei. The axon conduction velocity was determined to amount, on average, to 10.67 m/s for C-neurons and 15.84 m/s for L-neurons. In the vestibular nuclear complex, distribution of the fast and slow C- and L-neurons was studied. This study confirmed the previously made suggestion that C- and L-neurons of the frog, as sources of vestibular fibers, are distributed separately or, more often, as small groups, which leads to a patch-like somatotopy, rather than to formation of clearly separated fields.     

Effects of antidromic activity in gustatory nerve fibers on taste disc cells of the frog tongue

Summary Antidromic electrical stimulation of the lingual branch of the glossopharyngeal (IX) nerve of the frog was carried out while recording intracellular potentials of taste disc cells.Antidromic activation of sensory fibers resulted in depolarization of cells of the upper layer of the disc and most commonly in hyperpolarization of the cells in the lower layer. These changes in potential exhibited latencies greater than 1 s (Fig. 3), and thus cannot be due to electrotonic effects of action potentials in terminals of IX nerve fibers, which have much shorter conduction times. These cell potentials also showed summation, adaptation and post-stimulus rebound (Figs. 3, 4).Depending upon the chemical stimulus used, antidromic activity produced either depression or enhancement of gustatory fiber discharge in response to taste stimuli (Fig. 5).Alteration of the resting membrane potential by current injection did not significantly modify the antidromically evoked potentials (Fig. 8), whereas chemical stimulation of the tongue did (Fig. 7), indicating that these potential changes are not the result of passive electrical processes.These experimental results indicate that the membrane potential of taste disc cells can be modified by antidromic activity in their afferent nerves. This mechanism may be responsible for peripheral interactions among gustatory units of the frog tongue.The research was supported in part by NIH grant NS-09168.     

Further study of soma,dendrite, and axon excitation in single neurons

The present investigation continues a previous study in which the soma-dendrite system of sensory neurons was excited by stretch deformation of the peripheral dendrite portions. Recording was done with intracellular leads which were inserted into the cell soma while the neuron was activated orthodromically or antidromically. The analysis was also extended to axon conduction. Crayfish, Procambarus alleni (Faxon) and Orconectes virilis (Hagen), were used. 1. The size and time course of action potentials recorded from the soma-dendrite complex vary greatly with the level of the cell's membrane potential. The latter can be changed over a wide range by stretch deformation which sets up a "generator potential" in the distal portions of the dendrites. If a cell is at its resting unstretched equilibrium potential, antidromic stimulation through the axon causes an impulse which normally overshoots the resting potential and decays into an afternegativity of 15 to 20 msec. duration. The postspike negativity is not followed by an appreciable hyperpolarization (positive) phase. If the membrane potential is reduced to a new steady level a postspike positivity appears and increases linearly over a depolarization range of 12 to 20 mv. in various cells. At those levels the firing threshold of the cell for orthodromic discharges is generally reached. 2. The safety factor for conduction between axon and cell soma is reduced under three unrelated conditions, (a) During the recovery period (2 to 3 msec.) immediately following an impulse which has conducted fully over the cell soma, a second impulse may be delayed, may invade the soma partially, or may be blocked completely. (b) If progressive depolarization is produced by stretch, it leads to a reduction of impulse height and eventually to complete block of antidromic soma invasion, resembling cathodal block, (c) In some cells, when the normal membrane potential is within several millivolts of the relaxed resting state, an antidromic impulse may be blocked and may set up within the soma a local potential only. The local potential can sum with a second one or it may sum with potential changes set up in the dendrites, leading to complete invasion of the soma. Such antidromic invasion block can always be relieved by appropriate stretch which shifts the membrane potential out of the "blocking range" nearer to the soma firing level. During the afterpositivity of an impulse in a stretched cell the membrane potential may fall below or near the blocking range. During that period another impulse may be delayed or blocked. 3. Information regarding activity and conduction in dendrites has been obtained indirectly, mainly by analyzing the generator action under various conditions of stretch. The following conclusions have been reached: The large dendrite branches have similar properties to the cell body from which they arise and carry the same kind of impulses. In the finer distal filaments of even lightly depolarized dendrites, however, no axon type all-or-none conduction occurs since the generator potential persists to a varying degree during antidromic invasion of the cell. With the membrane potential at its resting level the dendrite terminals contribute to the prolonged impulse afternegativity of the soma. 4. Action potentials in impaled axons and in cell bodies have been compared. It is thought that normally the over-all duration of axon impulses is shorter. Local activity during reduction of the safety margin for conduction was studied. 5. An analysis was made of high frequency grouped discharges which occasionally arise in cells. They differ in many essential aspects from the regular discharges set up by the generator action. It is proposed that grouped discharges occur only when invasion of dendrites is not synchronous, due to a delay in excitation spread between soma and dendrites. Each impulse in a group is assumed to be caused by an impulse in at least one of the large dendrite branches. Depolarization of dendrites abolishes the grouped activity by facilitating invasion of the large dendrite branches.     

Model of spike propagation reliability along the myelinated axon corrupted by axonal intrinsic noise sources

We investigated how selected electromorphological parameters of myelinated axons influence the preservation of interspike intervals when the propagation of action potentials is corrupted by axonal intrinsic noise. Hereby we tried to determine how the intrinsic axonal noise influences the performance of axons serving as carriers for temporal coding. The strategy of this coding supposes that interspike intervals presented to higher order neurons would minimally be deprived of information included in interspike intervals at the axonal initial segment. Our experiments were conducted using a computer model of the myelinated axon constructed in a software environment GENESIS (GEneral NEural SImulation System). We varied the axonal diameter, myelin sheath thickness, axonal length, stimulation current and channel distribution to determine how these parameters influence the role of noise in spike propagation and hence in preserving the interspike intervals. Our results, expressed as the standard deviation of spike travel times, showed that by stimulating the axons with regular rectangular pulses the interspike intervals were preserved with a microsecond accuracy. Stimulating the axons with pulses imitating postsynaptic currents, greater changes of interspike intervals were found, but the influence of implemented noise on the jitter of interspike intervals was approximately the same.     

Simple models of stimulation of neurones in the brain by electric fields

The excitation of pyramidal cells in the motor cortex, produced by electric fields generated by distant electrodes or by electromagnetic induction, has been modelled. Linear, steady-state models of myelinated axons capture most of the geometrical aspects of neurone activation in electric fields. Some non-linear features can be approximated. Models with a proximal sealed-end and distal infinite axon, or of finite length, are both serviceable. Surface anodal stimulation produces hyperpolarisation of the proximal axon (closest to the anode) and depolarisation in the distal axon. The point of maximum depolarisation can be influenced by the location of the cathode (greater separation of anode and cathode causes more distal depolarisation). Axon bends can produce very localised depolarisation. Cathodal stimulation may be less effective than anodal as a result of anodal block of conduction of action potentials in the distal axon. The latencies of responses to anodal stimulation, recorded in the distal axon, will decrease as the stimulus strength is increased and the point of action potential initiation moves distally node by node. Larger jumps in latency will be produced when the point of action potential initiation moves from one axon bend to another.     

Long-range afferents in the rat spinal cord. 1. Numbers, distances and conduction velocities.

The caudal extent of the penetration of primary afferent axons from the T12 and L1 dorsal roots and sural nerve has been investigated in adult decerebrate spinal rats. Microelectrode stimulation at the root entry zone (REZ) and at further caudal points in the spinal cord was used to generate antidromic action potentials in single fibres recorded in dorsal roots or peripheral nerves. A total of 209 units were recorded in T12 and L1 dorsal roots and 27% of these could be antidromically activated 10 mm caudal to the REZ. Fifteen percent of the units could be stimulated at the L4-5 border, 15 mm caudal to the T12 segment whereas 4.5% of the axons could be stimulated 25 mm caudally in the S4 segment, 11 segments caudal to the entry segment. Similar recordings made from units in the sural nerve showed that of all the sural axons that penetrated to the L6 segment 50%, 18% and 2% of these reached the S1, S2 and S4 segments respectively. The conduction velocities of these units were clearly in the A-beta range when recorded in the nerve but decreased on entering the spinal cord and were reduced by 83% at their caudal end point. The results show that substantial numbers of primary afferents have long-ranging caudal branches in areas beyond the regions of known postsynaptic effects. The functions of these caudal projections are unclear but they may represent a potential substrate for the development of functional connections under conditions of disease or denervation.     

Propagation failure of action potentials at the bifurcation point of the slow axon innervating the extensor tibiae muscle ofDecticus albifrons (Orthoptera)

Summary Failure of conduction of nerve impulses has been observed at the bifurcation point of the metathoracic slow extensor tibiae motor axon (SETi) ofDecticus albifrons. Records from the region proximal and distal to the bifurcation point of the axon showed that during prolonged and repetitive stimulation and after a certain number of stimuli, proportional to the stimulating frequency, some SETi action potentials failed to cross this point (Fig. 1).Cross-sections of the metathoracic extensor motor nerve ofD. albifrons show that at the region of axonal bifurcation, both the neural lamella and the layer of glial cells (the sheath) around the SETi axons became thinner than the region proximal and distal to the bifurcation (Fig. 2).The possible role of the conduction block in the neuronal control of the muscle has been discussed.Abbreviations ETi extensor tibiae - SETi slow extensor tibiae - PE proximal electrode - DE distal electrode - SE stimulating electrode     

Model of spike propagation reliability along the myelinated axon corrupted by axonal intrinsic noise sources

We investigated how selected electromorphological parameters of myelinated axons influence the preservation of interspike intervals when the propagation of action potentials is corrupted by axonal intrinsic noise. Hereby we tried to determine how the intrinsic axonal noise influences the performance of axons serving as carriers for temporal coding. The strategy of this coding supposes that interspike intervals presented to higher order neurons would minimally be deprived of information included in interspike intervals at the axonal initial segment. Our experiments were conducted using a computer model of the myelinated axon constructed in a software environment GENESIS (GEneral NEural SImulation System). We varied the axonal diameter, myelin sheath thickness, axonal length, stimulation current and channel distribution to determine how these parameters influence the role of noise in spike propagation and hence in preserving the interspike intervals. Our results, expressed as the standard deviation of spike travel times, showed that by stimulating the axons with regular rectangular pulses the interspike intervals were preserved with a microsecond accuracy. Stimulation with pulses imitating postsynaptic currents, greater changes of interspike intervals were found, but the influence of implemented noise on the jitter of interspike intervals was approximately the same.     

Intermittency in the Hodgkin-Huxley model

We show that action potentials in the Hodgkin-Huxley neuron model result from a type I intermittency phenomenon that occurs in the proximity of a saddle-node bifurcation of limit cycles. For the Hodgkin-Huxley spatially extended model, describing propagation of action potential along axons, we show the existence of type I intermittency and a new type of chaotic intermittency, as well as space propagating regular and chaotic diffusion waves. Chaotic intermittency occurs in the transition from a turbulent regime to the resting regime of the transmembrane potential and is characterised by the existence of a sequence of action potential spikes occurring at irregular time intervals.