Yu. I. Arshavsky

Messages conveyed by spinocerebellar pathways during scratching in the cat. II. Activity of neurons of the ventral spinocerebellar tract

(1) The activity of neurons of the ventral spinocerebellar tract (VSCT) during scratching was studied in thalamic and decapitate cats. The neurons were identified antidromically either by stimulation of the hindlimb area in the anterior lobe of the cerebellum (in thalamic cats) or by stimulation of the contralateral ventrolateral funiculus of the spinal cord (in decapitate cats). The scratch reflex was elicited by stimulation of either the pinna (in thalamic cats) or the cervical spinal cord (in decapitate cats). In most experiments, animals were immobilized and the activity of VSCT neurons was recorded during fictitious scratching. (2) During both actual and fictitious scratching, the discharge of VSCT neurons was rhythmically modulated in relation with the scratch cycle: neurons fired in bursts separated with periods of silence. Phases of activity of different neurons were unevenly distributed over the scratch cycle: most neurons fired within the limits of the flexor phase of the cycle. (3) The firing pattern of VSCT neurons during fictitious scratching was similar to that during actual scratching. Therefore, rhythmical burst firing of VSCT neurons is determined mainly by central mechanisms and not by a rhythmical sensory input. (4) The firing pattern of VSCT neurons in decapitate cats was similar to that in thalamic cats. Therefore, rhythmical burst firing of VSCT neurons is determined mainly by the central spinal mechanism and not by supraspinal motor centers. (5) The VSCT neurons which fired in long bursts during the greater part of the flexor phase were usually activated during the latent period of scratching, while those firing later in the cycle were usually either inhibited or not affected during this period. (6) The antidromic response in most VSCT neurons could be evoked from a large number of points in the hindlimb area of the cerebellar anterior lobe, both in the vermis and in the pars intermedia. Due to such extensive branching of axons, each point of the cortex receives signals from neurons firing in different phases of the cycle. But axons of VSCT neurons firing in long bursts during the greater part of the flexor phase terminate more extensively in the pars intermedia, while axons of neurons firing later in the cycle terminate more extensively in the vermis. (7) The functioning of the VSCT is essentially similar to that of the spino-reticulocerebellar pathway (SRCP). Both pathways convey messages about activity of the central spinal mechanism generating the motor output pattern of scratching, but the VSCT is active mainly in the flexor phase of the scratch cycle and the SRCP in the extensor one. A hypothesis is advanced that these pathways monitor activity of different groups of spinal interneurons.

Control of locomotion in marine mollusc Clione limacina. I. Efferent activity during actual and fictitious swimming.

Summary1.The marine mollusc Clione limacina swims by making rhythmic movements (with a frequency of 1–5 Hz) of its two wings. Filming demonstrated that the wings perform oscillatory movements in the frontal plane of the animal. During both the upward and downward movements of the wing, its posterior edge lagged behind the anterior one, i.e. the wing plane was inclined in relation to the longitudinal axis of an animal. As a result of this inclination, the wing oscillations in the frontal plane produce a force directed forwards. 2.In restrained animals with the body cavity opened (a whole-animal preparation), the wing position, electrical activity in the wing nerve and activity of two identified efferent neurons (1A and 2A) were recorded during locomotory wing movements. There were two bursts of activity in the wing nerve during the locomotory cycle, the first one corresponding to the excitation of efferent neurons controlling the wing elevation, and the second one, to the excitation of efferent neurons controlling the lowering of the wing. Neurons 1A and 2A fired reciprocally at the beginning of the phase of elevating and lowering the wing, respectively. During excitation of one of the neurons, an IPSP appeared in its antagonist. 3. A pair of isolated pedal ganglia of Clione was capable of generating the locomotory rhythm (“fictitious swimming”). In fictitious swimming, as in actual swimming, there were two bursts of activity in the wing nerve per locomotory cycle, and the 1A and 2A neurons fired reciprocally. Homologous neurons from the left and right ganglia fired inphase. A single pedal ganglion was also capable of generating the locomotory rhythm. 4.Serotonin (10-5–10-6 M) increased the locomotor activity both in the whole-animal preparation and in the isolated pedal ganglia.

Control of locomotion in marine mollusc Clione limacina II. Rhythmic neurons of pedal ganglia

Summary1.Activity from neurons in isolated pedal ganglia of Clione limacina was recorded intracellularly during generation of rhythmic swimming. To map the distribution of cells in a ganglion, one of two microelectrodes was used to monitor activity of the identified neuron (1A or 2A), while the second electrode was used to penetrate successively all the visible neurons within a definite area of the ganglion. In addition, pairs of neurons of various types were recorded in different combinations with each other. Intracellular staining of neurons was also performed. 2.Each ganglion contained about 400 neurons, of which about 60 neurons exhibited rhythmic activity related to a swim cycle. These rhythmic neurons were divided into 9 groups (types) according to axonal projections, electrical properties and the phase of activity in a swim cycle. Three types of interneurons and six types of efferent neurons were distinguished. 3.Type 7 and 8 interneurons generated only one spike of long (50–150 ms) duration per swim cycle. Type 7 interneurons discharged in the phase of the cycle that corresponded (in actual swimming) to the dorsal movement of wings (D-phase). Type 8 interneurons discharged in the opposite phase corresponding to the ventral movement of wings (V-phase). With excitation of type 7 interneurons, an IPSP appeared in the type 8 interneurons, and vice versa. Neuropilar branching of these neurons was observed in the ipsilateral ganglion. In addition, they sent an axon to the contralateral ganglion across the pedal commissure. 4.Efferent neurons (i.e. the cell sending axons into the wing nerve) generated spikes of 1–5 ms duration. Type 1 and 3 neurons were excited in the D-phase of a swim cycle and were inhibited in the V-phase. Type 2 and 4 neurons were excited in the V-phase and inhibited in the D-phase. Type 10 neurons received only an excitatory input in the V-phase, while type 6 neurons received only an inhibitory input in the D-phase. 5. Type 12 interneurons were non-spiking cells, they generated a stable depolarization (“plateau”) throughout most of the V-phase. 6. Neurons of the same type from one ganglion (except for type 6) were electrically coupled to each other. There were also electrical connections between most neurons firing in the same phase of the cycle, i.e. between types 3 and 7, as well as between types 2, 4 and 8. Type 7 interneurons from the left and right ganglia were electrically coupled, the same was true for type 8 interneurons.

Control of locomotion in marine mollusc Clione limacina III. On the origin of locomotory rhythm

Summary1. Neurons from the isolated pedal ganglia of the marine mollusc Clione limacina were recorded from intracellularly during generation of the locomotory rhythm. Polarization of single type 7 or type 8 interneurons (which discharge in the D-and V-phases of a swim cycle, respectively) strongly affected activity of the rhythm generator. Injection of depolarizing and hyperpolarizing current usually resulted in shortening and lengthening of a swim cycle, respectively. A short pulse of hyperpolarizing current shifted the phase of the rhythmic generator. The same effect could be evoked by polarization of efferent neurons of types 2, 3 and 4 which are electrically coupled to interneurons. On the contrary, polarization of types 1, 6 and 10 efferent neurons, having no electrical connections with interneurons, did not affect the locomotory rhythm. 2. A number of observations indicate that type 7 and 8 interneurons constitute the main source of postsynaptic potentials that were observed in all the “rhythmic” neurons of the pedal ganglia. Type 7 interneurons excited the D-phase neurons and inhibited the V-phase neurons; type 8 interneurons produced opposite effects. 3. Tetrodotoxin eliminated spike generation in all efferent neurons of the pedal ganglia, while in interneurons spike generation persisted. After blocking the spike discharges in all the efferent neurons, type 7 and 8 interneurons were capable of generating alternating activity. One may conclude that these interneurons determine the main features of the swim pattern, i.e., the rhythmic alternating activity of two (D and V) populations of neurons. 4. Both type 7 and type 8 interneurons were capable of endogenous rhythmic discharges with a period like that in normal swimming. This was demonstrated in experiments in which one of the two populations of “rhythmic” neurons (D or V) was inhibited by means of strong electrical hyperpolarization, as well as in experiments in which interaction between the two populations, mediated by chemical synapses, was blocked by Co2+ ions. 5. Type 7 and 8 interneurons were capable of “rebound”, i.e. they had a tendency to discharge after termination of inhibition. 6. V-phase neurons exerted not only inhibitory but also excitatory action upon D-phase neurons, the excitatory action being longer than the inhibitory one. 7. The main experimental findings correspond well to the model of rhythm generator consisting of two half centres possessing endogenous rhythmic activity. The half-centres exert strong, short duration inhibitory and weak long duration excitatory actions upon one another. The behaviour of such a model is considered and compared with that of the locomotor generator of Clione.

Activity of neurons of cerebellar nuclei during fictitious scratch reflex in the cat. II. Interpositus and lateral nuclei

The activity of neurons of the interpositus and lateral cerebellar nuclei was recorded during fictitious scratch reflex in thalamic cats immobilized with Flaxedil. Interpositus neurons were identified by antidromic response to stimulation of the contralateral red nucleus. The interpositus neurons responding to passive movements of the ipsilateral hindlimb manifested rhythmical modulation of the discharge in relation with the scratch cycle. The neurons generated bursts of impulses separated by periods of silence. Different neurons were active in different parts of the scratch cycle, but most of them were active in the second half of the flexor phase. When the scratch reflex was evoked on the contralateral side, rhythmical modulation was observed in about half of the neurons, and it was less pronounced than in the case of ipsilateral scratching. Rhythmical modulation of cerebellar neurons during fictitious scratching is determined by signals coming from the central spinal mechanism, generating rhythmical oscillations, via the ventral spinocerebellar tract (VSCT) and the spinoreticulocerebellar pathway (SRCP). Results of separate transections of these pathways showed that the VSCT plays the crucial role in modulating interpositus neurons. Neurons of the lateral nucleus exhibited no modulation during fictitious scratching.

Messages conveyed by spinocerebellar pathways during scratching in the cat. I. Activity of neurons of the lateral reticular nucleus

(1) Signals transmitted to the cerebellum by the spino-reticulocerebellar pathway (SRCP) during scratching were studied. For this purpose, the activity of neurons of the lateral reticular nucleus (LRN), which are the last-order neurons of the SRCP, was recorded during scratching in thalamic cats. Scratching was evoked by stimulation of the pinna. LRN neurons were identified antidromically by stimulation of the hindlimb area in the cerebellar anterior lobe. In most experiments, animals were immobilized with Flaxedil, and stimulation of the pinna resulted in fictitious scratching, i.e., in periodical reciprocal activity of flexor and extensor motoneurons typical of actual scratching. (2) During both actual and fictitious scratching, the discharge frequency of LRN neurons was rhythmically modulated in relation with the scratch cycle. Most LRN neurons fired in short high-frequency bursts of spikes which coincided (completely or partly) with the extensor phase of the cycle. In this respect the SRCP differs from the ventral spinocerebellar tract (VSCT) which is maximally active in the flexor phase of the cycle. (3) The firing pattern of LRN neurons during fictitious scratching was similar to that during actual scratching. Therefore, the rhythmical burst firing of LRN neurons is determined mainly by the central mechanisms and not by the rhythmical sensory input. (4) Rhythmical modulation of LRN neurons disappeared after transection of the ipsilateral lateral funiculus of the spinal cord in which spinoreticular fibers are located. On the other hand, considerable reduction of rhythmical activity in descending brainstem-spinal pathways after contralateral hemisection of the spinal cord did not affect the discharge pattern of LRN neurons. These two facts indicate that the SRCP conveys mainly messages about activity of the central spinal mechanisms, and that influences of supraspinal motor centers.on LRN neurons and on spinoreticular neurons are of minor importance. (5) Axonal terminations of LRN neurons are distributed rather evenly over the hindlimb area in the anterior lobe of the cerebellum. Therefore, messages about the events, which happen within the spinal cord in the vicinity of the extensor phase of the cycle, arrive at every point of the hindlimb area.

Activity of neurons of cerebellar nuclei during fictitious scratch reflex in the cat. I. Fastigial nucleus

The activity of neurons located in the rostral part of the fastigial nucleus was recorded during the fictitious scratch reflex in thalamic cats immobilized with Flaxedil. The activity of most of the neurons was rhythmically modulated in relation to the scratch cycle: they generated bursts of impulses separated by periods of silence or, in a few cases, by periods of reduced activity. The discharge pattern of neurons was usually the same during both ipsilateral and contralateral scratch reflex: in both cases their maximum activity could be observed in the extensor phase of the scratch cycle. Rhythmical modulation of fastigial neurons is determined by signals coming from the central spinal mechanism, generating rhythmical oscillations, via the ventral spinocerebellar tract (VSCT) and the spinoreticulo-cerebellar pathway (SRCP). Results of separate transections of these pathways suggest that the SRCP determines the periodical excitationm of fastigial neurons and the VSCT their periodical inhibition.

Messages conveyed by descending tracts during scratching in the cat. II. Activity of rubrospinal neurons

(1) The activity of vestibulospinal (VS) neurons giving axons to the lumbosacral spinal cord was recorded during scratching in thalamic and decerebrate cats. The most part of the experiments was carried out on curarized cats, in which fictitious scratching13, i.e. rhythmical activity of motoneurons typical of actual scratching, was evoked. (2) During both actual and fictitious scratching, the discharge frequency of many VS neurons was rhythmically modulated in relation to the scratch cycle. Most modulated neurons were maximally active in the extensor phase of the cycle. (3) The firing pattern of VS neurons during fictitious scratching was similar to that during actual scratching. Therefore, rhythmical modulation of VS neurons is determined mainly by central mechanisms and not be a rhythmical sensory input. (4) In decerebellate cats, rhythmical modulation was not found during either actual or fictitious scratching. (5) Transection of the ventral spinocerebellar tract (VSCT) resulted in considerable reduction of rhythmical modulation of VS neurons during fictitious scratching, while transection of the spino-reticulocerebellar pathway (SRCP) resulted in just a small decrease of modulation. Therefore, of the two pathways (VSCT and SRCP) transmitting messages about intraspinal processes to the cerebellum during scratching6,7, the VSCT is of major importance for modulating VS neurons.

Control of locomotion in marine mollusc Clione limacina. VI. Activity of isolated neurons of pedal ganglia.

SummaryIn the pteropodial mollusc Clione limacina, the rhythmic locomotor wing movements are controlled by the pedal ganglia. The locomotor rhythm is generated by two groups of interneurons (groups 7 and 8) which drive efferent neurons. In the present paper, the activity of isolated neurons, which were extracted from the pedal ganglia by means of an intracellular electrode, is described. The following results have been obtained: 1. Isolated type 7 and 8 interneurons preserved the capability for generation of prolonged (100–200 ms) action potentials. The frequency of these spontaneous discharges was usually within the limit of locomotor frequencies (0.5–5 Hz). By de- or hyperpolarizing a cell, one could usually cover the whole range of locomotor frequencies. This finding demonstrates that the locomotor rhythm is indeed determined by the endogenous rhythmic activity of type 7 and 8 interneurons. 2. Type 1 and 2 efferent neurons, before isolation, could generate single spikes as well as high-frequency bursts of spikes. These two modes of activity were also observed after isolating the cells. Thus, the bursting activity of type 1 and 2 neurons, demonstrated during locomotion, is determined by their own properties. Type 3 and 4 efferent neurons generated only repeated single spikes both before and after isolation. 3. The activity of the isolated axons of type 1 and 2 neurons did not differ meaningfully from the activity of the whole cells. Furthermore, in the isolated pedal commissure, we found units whose activity (rhythmically repeating prolonged action potentials) resembled the activity of type 7 and 8 interneurons. These units seemed to be the axons of type 7 and 8 interneurons. Thus, different parts of the cell membrane (soma and axons) have similar electric properties.

Control of locomotion in marine mollusc Clione limacina IV. Role of type 12 interneurons

Summary1. Type 12 interneurons in pedal ganglia of Clione limacina exerted a strong influence upon the locomotor generator during “intense” swimming. These neurons generated “plateau” potentials, i.e. their membrane potential had two stable states: the “upper” one when a neuron was depolarized, and the “down” one, separated by 30–40 mV. The interneurons could remain in each state for a long time. Short depolarizing and hyperpolarizing current pulses, as well as excitatory and inhibitory postsynaptic potentials, could transfer the interneurons from one state to another. 2. When the pedal ganglia generated the locomotory rhythm, type 12 neurons received an EPSP and passed to the “upper” state in the V2-phase of a locomotor cycle. They remained at this state until the beginning of the D1-phase when they received an IPSP and passed to the “down” state. The EPSP in type 12 neurons was produced by type 8d neurons, and the IPSP by type 7 neurons. 3. Type 12 neurons exerted inhibitory influences upon many neurons active in the V1 and V2 phases, and excitatory influences upon the D-phase interneurons (type 7). 4. The functional role of type 12 neurons was to limit the activity of neurons discharging in the V-phase of a locomotory cycle. In addition, they enhanced the excitation of the D-phase neurons and promoted, thus, the transition from the V-phase to the D-phase.