R. Llinás

1. Intradendritic recordings from Purkinje cells in vitro indicate that white matter stimulation produces large synaptic responses by the activation of the climbing fibre afferent, but antidromic potentials do not actively invade the dendritic tree. 2. Climbing fibre responses may be reversed in a manner similar to that observed at the somatic level. However, the reversal does not show the biphasicity often seen at somatic level. 3. Input resistance of these dendrites was found to range from 15 to 30 M omega. The non-linear properties seen at the somatic level for depolarizing currents are also encountered here. However, there seems to be less anomalous rectification. 4. Detailed analysis of repetitive firing of Purkinje cells elicited by outward DC current shows that, as in the case of the antidromic invasion, the fast somatic potentials (s.s.) do not invade the dendrite actively. However, the dendritic spike bursts (d.s.b.s) interposed between the s.s. potentials are most prominent at dendritic level. 5. Two types of voltage-dependent Ca responses were observed. At low stimulus level a plateau-like depolarization is accompanied by a prominent conductance change; further depolarization produces large dendritic action potentials. These two classes of response are TTX-resistant but are blocked by Cd, Co, Mn or D600, or by the removal of extracellular Ca. 6. Following blockage of the Ca conductance, plateau potentials produced by a non-inactivating Na conductance are observed mainly near the soma indicating that this voltage-dependent conductance is probably associated with the somatic membrane. 7. Spontaneous firing in Purkinje cell dendrites is very similar to that observed at the soma. However, the amplitude of these bursts is larger at dendritic level. It is further concluded that these TTX-insensitive spikes are generated at multiple sites along the dendritic tree. 8. Six ionic conductances seem to be involved in Purkinje cell electroresponsiveness: (a) an inactivating and (b) a non-inactivating Na conductance at or near the soma, (c) a spike- and (d) a plateau-generating Ca conductance, and (e) voltage-dependent and (f) Ca-dependent K currents. 9. The possible role of these conductances in Purkinje cell integration is discussed.

The electroresponsive properties of guinea-pig thalamic neurones were studied using an in vitro slice preparation. A total of 650 cells were recorded intracellularly comprising all regions of the thalamus; of these 229 fulfilled our criterion for recording stability and were used as the data base for this report. The resting membrane potential for thirty-four representative neurones which were analysed in detail was -64 +/- 5 mV (mean +/- S.D.), input resistance 42 +/- 18 M omega, and action potential amplitude 80 +/- 7 mV. Intracellular staining with horseradish peroxidase and Lucifer Yellow revealed that the recorded cells had different morphology. In some their axonal trajectory characterized them as thalamo-cortical relay cells. Two main types of neuronal firing were observed. From a membrane potential negative to -60 mV, anti- or orthodromic and direct activation generated a single burst of spikes, consisting of a low-threshold spike (l.t.s.) of low amplitude and a set of fast superimposed spikes. Tonic repetitive firing was observed if the neurones were activated from a more positive membrane potential; this was a constant finding in all but two of the cells which fulfilled the stability criteria. The l.t.s. response was totally inactivated at membrane potentials positive to -55 mV. As the membrane was hyperpolarized from this level the amplitude of the l.t.s. increased and became fully developed at potentials negative to -70 mV. This increase is due to a de-inactivation of the ionic conductance generating this response. After activation the l.t.s. showed refractoriness for approximately 170 ms. Deinactivation of l.t.s. is a voltage- and time-dependent process; full de-inactivation after a step hyperpolarization to maximal l.t.s. amplitude (-75 to -80 mV) requires 150-180 ms. Membrane depolarization positive to -55 mV generated sudden sustained depolarizing 'plateau potentials', capable of supporting repetitive firing (each action potential being followed by a marked after-hyperpolarization, a.h.p.). The a.h.p. and the plateau potential controlled the voltage trajectory during the interspike interval and, with the fast spike, constitute a functional state where the thalamic neurone displayed oscillatory properties. Frequency-current (f-I) plots from different initial levels of membrane potential were obtained by the application of square current pulses of long duration (2s). From resting membrane potential and from hyperpolarized levels a rather stereotyped onset firing rate was observed due to the presence of the l.t.s.(ABSTRACT TRUNCATED AT 400 WORDS)

The ionic requirements for electro-responsiveness in thalamic neurones were studied using in vitro slice preparations of the guinea-pig diencephalon. Analysis of the current-voltage relationship in these neurones revealed delayed and anomalous rectification. Substitution of Na+ with choline in the bath or addition of tetrodotoxin (TTX) abolished the fast spikes and the plateau potentials, described in the accompanying paper. Ca2+ conductance blockage with Co2+, Cd2+ or Mn2+, or replacement of Ca2+ by Mg2+ abolished the low-threshold spikes (l.t.s.). Substitution with Ba2+ did not significantly increase the duration of the l.t.s., suggesting that under normal conditions the falling phase of this response is brought about by inactivation of the Ca2+ conductance. The after-hyperpolarization (a.h.p.) following fast spikes was markedly reduced in amplitude and duration by bath application of Cd2+, Co2+ or Mn2+, indicating that a large component of this response is generated by a Ca2+-dependent K+ conductance (gK[Ca]). Following hyperpolarizing current pulses, the membrane potential showed a delayed return to base line. This delay is produced by a transient K+ conductance as it can be modified by changing the drive force for K+. Presumptive intra-dendritic recording demonstrated high-threshold Ca2+ spikes (h.t.s.s.) which activate a gK[Ca]. Such h.t.s.s. were also seen at the somatic level when K+ conductance was blocked with 4-aminopyridine. It is proposed that the intrinsic biophysical properties of thalamic neurones allow them to serve as relay systems and as single cell oscillators at two distinct frequencies, 9-10 and 5-6 Hz. These frequencies coincide with the alpha and theta rhythms of the e.e.g. and, in the latter case, with the frequency of Parkinson's tremor.

1. The electrical activity of Purkinje cells was studied in guinea-pig cerebellar slices in vitro. Intracellular recordings from Purkinje cell somata were obtained under direct vision, and antidromic, synaptic and direct electroresponsiveness was demonstrated. Synaptic potentials produced by the activation of the climbing fibre afferent could be reversed by direct membrane depolarization. 2. Input resistance of impaled neurones ranged from 10 to 19 M omega and demonstrated non-linearities in both hyperpolarizing and depolarizing directions. 3. Direct activation of a Purkinje cell indicated that repetitive firing of fast somatic spikes (s.s.) occurs, after a threshold, with a minimum spike frequency of about 30 spikes/sec, resembling the '2-class' response of crab nerve (Hodgkin, 1948). 4. As the amplitude of the stimulus was increased, a second form of electroresponsiveness characterized by depolarizing spike bursts (d.s.b.) was observed and was often accomppanied by momentary inactivation of the s.s. potentials. Upon application of tetrodotoxin (TTX) or removal of Na+ ions from the superfusion fluid, the s.s. potentials were abolished while the burst responses remained intact. However, Ca conductance blockers such as Co, Cd, Mn and D600, or the replacement of Ca by Mg, completely abolish d.s.b.s. 5. If Ca conductance was blocked, or Ca removed from the superfusion fluid without blockage of Na conductance, two types of Na-dependent electroresponsiveness were seen: (a) the s.s. potentials and (b) slow rising all-or-none responses which reached plateau at approximately -15 mV and could last for several seconds. These all-or-none Na-dependent plateau depolarizations outlasted the stimulus and were accompanied by a large increase in membrane conductance. Within certain limits the rate of rise and amplitude of the plateau were independent of stimulus strength. The latency, however, was shortened as stimulus amplitude was increased. These potentials were blocked by TTX or by Na-free solutions. 6. Substitution of extracellular Ca by Ba or intracellular injection of tetraethylammonium generated prolonged action potentials lasting for several seconds and showing a plateau more ositive than those obtained in norrmal circumstances by either non-inactivating Na or Ca currents. 7. Spontaneous firing of the Purkinje cell was characterized by burst-like activity consisting of both s.s. and d.s.b. responses. Addition of TTX to the bath left the basic spontaneous activity and its frequency unaltered, indicating tha Ca spiking and Ca-dependent K conductance changes are the main events underlying this oscillatory behaviour. 8...

1. A single climbing fibre makes an extraordinarily extensive synaptic contact with the dendrites of a Purkinje cell. Investigation of this synaptic mechanism in the cerebellum of the cat has been based on the discovery by Szentagothai & Rajkovits (1959) that the climbing fibres have their cells of origin in the contralateral inferior olive.2. Stimulation in the accessory olive selectively excites fibres that have a powerful synaptic excitatory action on Purkinje cells in the contralateral vermis, evoking a repetitive spike discharge of 5-7 msec duration. Almost invariably this response had an all-or-nothing character. In every respect it corresponds with the synaptic action that is to be expected from climbing fibres.3. Intracellular recording from Purkinje cells reveals that this climbing fibre stimulation evokes a large unitary depolarization with an initial spike and later partial spike responses superimposed on a sustained depolarization.4. Typical climbing fibre responses can be excited, but in a much less selective manner, by stimulation of the olive-cerebellar pathway in the region of the fastigial nucleus, there being often a preceding antidromic spike potential of the Purkinje cell under observation.5. Impaled Purkinje cells rapidly deteriorate with loss of all spike discharge, the climbing fibre response being then reduced to an excitatory post-synaptic potential. This potential shows that stimulation of the inferior olive may evoke two or more discharges at about 2 msec intervals in the same climbing fibre. The complexity of neuronal connexions in the inferior olive is also indicated by the considerable latency range in responses.6. A further complication is that, with stimulation in the region of the fastigial nucleus, the initial direct climbing fibre response is often followed by a reflex discharge, presumably from the inferior olive, which resembles the responses produced by inferior olive stimulation in being often repetitive.7. Typical climbing fibre responses have been evoked by peripheral nerve stimulation and frequently occur spontaneously.8. An account is given of the way in which the responses evoked by climbing fibres in the individual Purkinje cells can account for the potential fields that an inferior olive stimulus evokes on the surface and through the depth of the cerebellar cortex.9. By the application of currents through the recording intracellular electrode it has been possible to effect large changes in the excitatory post-synaptic potential produced by a climbing fibre, it being diminished and even reversed with depolarizing currents and greatly increased by hyperpolarizing currents.