Acta physiol. scand. 1975.95. 431-440 From the Department of Physiology, University of Umel, Sweden
Rubrospinal Control of Static and Dynamic Fusimotor Neurones BY
B. APPELBERG, T. JENESKOG and H. JOHANSSON Received 7 May 1975
Abstract APPELBERG, B., T. JENESKOG and H. JOHANSSON. Rubrospinal control of static and dynamic fusimotor neurones. Acta physiol. scand. 1975. 95. 431-440. Rubrospinal effects on about 60 extracellularly recorded y-motoneurones were studied in anesthetized cats. All cells were antidromically identified from various muscle nerves. 23 cells were regarded as dynamic as they were activated from a mesencephalic region previously known to influence selectively muscle spindle dynamic sensitivity. The pattern of rubrospinal influence on static fusimotor neurones to different muscles closely followed that previously demonstrated for a-motoneurones with predominantly excitation of flexor neurones and excitation or inhibition in equal amounts of extensor cells. Dynamic fusimotor neurones were influenced in a strictly reciprocal manner with excitation of flexor cells and inhibition of extensor cells except for a few neurones which could not be reached from nucleus ruber. Evidence was also obtained indicating that the shortest path from nucleus ruber to static fusimotor neurones involves one interneurone.
The effect of electrical stimulation within the red nucleus on the fusimotor system was studied by Appelberg (1962) and by Appelberg and Kosary (1963) by the indirect method of recording from muscle spindle afferents. It was regularly observed that such stimulation caused inhibition of spontaneous activity in extensor spindles while flexor spindles were instead excited. These findings were in general agreement with the notion at that time that the red nucleus was excitatory to flexor and inhibitory to extensor muscles (Sasaki, Namikawa and Hashiramoto 1960). In recent years the control function of the rubrospinal tract as well as of other systems descending from the mesencephalon and acting on the neuronal machinery in lumbar segments have been extensively and carefully studied by Lundberg and coworkers (cf. Hongo, Jankowska and Lundberg 1969 a, b; 1972 a, b and Baldissera, Lundberg and Udo 1972 a, b). These studies have, among other things, given extremely detailed information about rubrospinal control of hind limb a-motoneurones and also carefully analyzed methodological aspects of importance when using electrical stimulation in the rubral region. With this in mind and considering current theories of linkage between a- and y-motoneurones in motor control it seemed of great interest to reinvestigate in greater detail and 43 1
432
B. APPELBERG, T. JENESKOG AND H. JOHANSSON
with another technique the matter of rubrospinal control of hind limb fusimotor neurones. A detailed knowledge of a descending system selectively influencing dynamic fusimotor neurones (cf. Appelberg and Jeneskog 1972, Jeneskog 1974 a) was utilized as a methodological tool allowing the separation of y-motoneurones into the two functional groups, static and dynamic.
Methods The results to be presented were obtained in experiments on 18 cats. Of these 11 were anesthetized with halothane (usually 1.0-1.5%) administered in a mixture of Oxygen (1/3) and Nitrous Oxide (2/3), 4 with a-Chloralose (60 mg/kg) administered intravenously after induction with halothane and 3 with pentobarbital (40 mg/kg) given intraperitoneally with later supplementary intravenous doses. Essentially similar findings were obtained regardless of type of anesthesia. Blood pressure and expiratory CO, as well as rectal temperature and temperatures in the paraffin pools were continuously monitored throughout the experiments.
Operation The operative procedure included a low thoracic and a lumbar laminectomy. Th e dorsal columns and the right spinal half were sectioned in Th12-13. The lumbar spinal cord was prepared for microelectrode recording from himd limb motoneurone pools. In the left hind limb the following nerves were dissected and mounted for stimulation: posterior bicepssemitendinosus (PBSt), anterior biceps-semimembranosus (ABSm), gastrocnemius-soleus (GS),plantaris (P),flexor digitorum and hallucis longus (FDL), deep peroneal (DP), superficial peroneal (SP) and tibia1 (Tib). In one experiment the tenuissimus nerve was prepared instead and later used for dissection of filaments (see Results, Fig. 3). The sensorimotor cortex was acutely ablated bilaterally in all experiments. Access to the right mesencephalon for stimulating electrodes and to the left paramedian lobule of the cerebellum for surface recording was arranged.
Recording Extracellular recording was made from y-motoneurones with glass capillary electrodes filled with 4 M NaCI. The impedance of the electrodes was usually about 1-2 M a . Only cells which could be antidromically invaded from one of the dissected muscle nerves and thus identified were accepted. The conduction velocities of the axons could then be determined by also measuring the conduction distance along the different nerves after each experiment. Only cells having a spontaneous activity could be tested with regard to inhibitory influence from the brain, and therefore each cell included in the material exhibited such an activity during at least part of the recording session. Surface recording of the rubrospinal tract volley on the dorsolaterd aspect of the cord in the exposed thoracic segment was employed to guide the placement of a stimulating electrode in the red nucleus (NR). Surface recording of climbing fibre responses in the D-zone of the cerebellar paramedian lobule (PM) was employed when placing a stimulating electrode in a mesencephalic area previously shown to selectively influence dynamic fusimotor neurones (MesADC-cf. Appelberg 1967 and Jeneskog 1974 a and Discussion).
Stimulation For stimulation in the brain two medio-lateral grids of 3 glasscovered platinum-iridium electrodes with 1 mm interelectrode distance were used. One of the grids was angled 15" to the vertical plane as all electrodes had to be positioned at approximately the same rostro-caudal level but at different depths. A presumably selective activation of the rubro-spinal tract was obtained by stimulation of interpositorubral fibres in the caudo-ventral part of or just ventrally to the caudal part of the red nucleus. The correct placement of the stimulating electrode was aided by observing the descending volley change from a short latency direct response to a transsynaptic response with longer latency when slowly proceeding with the electrode through the red nucleus (cf. Baldissera et al. 1972 b). This placement allowed a maximal activation of the rubro-spinal tract at low stimulating strength at a considerable distance away from the MesADC.
433
NUCLEUS RUBER ON FUSIMOTOR NEURONES
P-Y 31 rn/s
NR90pA
C
B
-
E F
NR 3 0 p A
I
I
MesADC 6 0 p A
o.5s
Fig. 2
Fig. I
Fig. 1. A and B are camera lucida drawings of transverse sections through the mesencephalon at a rubral level (B about I mm caudal to A) with stimulating sites indicated. Three stimulating shocks at 600 Hz at the site in A activates rubrospinal tract as revealed by surface recording from the cord in C (upper trace). The stimulating point in B causes no rubrospinal volley (D, upper trace) but instead a climbing fibre response in the paramedian lobule of the cerebellum is evoked (D, lower trace). Positivity is recorded upwards in this and all subsequent figures. Time calibration 5 ms, voltage calibration 100 pV to upper traces in C and D, 200 ,uV to lower traces. Fig. 2. A, B, identification of plantaris fusimotor neurone (see text). C , descending volley in Th,, to N R stimulation. D, PM-response evoked from MesADC (upper trace). Note lack of descending volley (lower trace). Time calibration below B is 10 ms to A, B and D, 4 ms to C . Voltage calibration is IOOyV for upper trace in D. E, inhibition of spontaneous fusimotor activity by repetitive stimulation in N R during time indicated by black line below record. F, activation of same fusimotor neurone from MesADC.
An equally selective activation of the descending pathway to dynamic fusimotor neurones was attempted by placing a second stimulating electrode dorsally in the low-threshold region for a D-zone climbing fibre response in the PM. A direct activation of rubrospinal cells could usually be obtained from this region, but only at high stimulating intensities. Only effects on y-motoneurones at strengths well below the threshold for NR-activation were considered. Also at such lower stimulating strengths a descending volley could sometimes be recorded in the thoracic segment during train stimulation. This volley was similar to that described by Baldissera el al. (1972 a) as being characteristic for the dorsal reticulo-spinal system. In Fig. I the position of stimulating sites for rubro-spinal activation (A) and within the MesADC (B) in one experiment, where histological control was made, is illustrated. In C and D the physiological criteria used for correct placement in all experiments are illustrated. In a few early experiments only NR-stimulation took place and of the total cell-material of 63 cells 13 were thus not tested from MesADC (see under Results).
Results 1. Effects from N R and MesADC on y-motoneurones
Table I summarizes the results obtained in one experiment (same experiment as Fig. 1). Nine y-motoneurones were recorded from, 6 of which were completely tested with regard to 33 - 755882
434
B. APPELBERG, T. JENESKOG AND H. JOHANSSON
TABLE 1. Effects from nucleus ruber (NR) and mesencephalic area for dynamic control of muscle spindles (MesADC) on nine fusirnotor neurones in 1 expt. Thresholds for effects given in PA. t actrvation, - inhibition, O=no influence. ~
~
Cell no. ~~
~~
Type
1
FDL
DP DP DP DP DP DP PBSt FDL
6
NR
PA
-
50 50 60 30 40 40 70 60
MesADC pA
~~
2 3 4 5 7 8 9
-
~~~
Cond. vel. 32 m/s 32 rn/s 21 m/s 34 m/s 21 m/s 30 rn/s 23 m/s 16mjs
+ +
+ i+ -t+ -
+ + + + +
0
60 70 70 60 70 40
0
Threshold for rubrospinal volley from NR 30 PA. Threshold for rubrospinal volley from MesADC 90 / L A . Threshold for PM-response from MesADC 30 /LA.
effects evoked from the two central regions, while 3 cells were kept only for short periods of time and not fully tested. The threshold for evoking a rubrospinal volley from the electrode just ventrally to the N R was about 30 PA. From the electrode placed in MesADC a weak rubrospinal volley could be evoked at stimulating strengths above 90 PA. A PM-response was, on the other hand, evoked already at 30 pA. It is clear from the Table that all effects on the different cells were obtained at stimulating intensities which makes spread of current from one region to the other unlikely. This conclusion is supported by the fact that cell nr I , which was reciprocally influenced from the two regions at low stimulating intensities, was controlled in the same way, but stronger, when stimulation was increased to 90 ,uA in N R and to 80 pA in MesADC. The central effects on all cells included in the material presented have been studied under similar circumstances. A cell influenced in the same reciprocal way as the FDL-y (cell 1) in Table I but belonging to the plantaris muscle is illustrated in Fig. 2. This cell was slightly difficult to identify as it responded antidromically from the nerve together with another cell (A). From the recordings in A and B it appears, however, that when a spontaneous spike of the cell occurred closely enough to the antidromic stimulus (B) a collision of orthodromic and antidromic spikes took place in the axon and the antidromic response disappeared. The remaining smaller cell belonging to the same nerve was not spontaneously active, nor was it influenced by the central stimuli. The rubrospinal volley evoked from the NR-electrode is illustrated in C , and the characteristic PM-response from the MesADC-electrode in D (note lack of volley in Th 13). In Fig. 2 E is seen a strong inhibitory influence on the spontaneously active cell from NR at low stimulus intensity. In F the cell is instead shown to be strongly activated from MesADC. In this experiment the thresholds for detectable descending rubrospinal volley from the NR-electrode and for PM-response from the MesADC-electrode were both around 40 [LA.A weak rubrospinal volley for MesADC-stimulation was seen at 100 PA. In Table 11 the whole material of 63 cells and the effects from MesADC and NR is presented. The cells have been divided into 3 subpopulations on the basis of the effects upon them from MesADC. As previous work on muscle spindle afferents has revealed that the
435
NUCLEUS RUBER ON FUSIMOTOR NEURONES
TABLE 11. Effects from NR and MesADC on whole material of 63 cells. A. Cells not influenced from MesADC (Static cells) Effect from N R GS FDL P ABSm
PBSt
DP
Activation Inhibition
5
6
0
I
0
5
13
14
Extensor
Flexor
Number
0
1
5
0 1
1
0
9 0
10
1
1
I
9
5
B. Cells influenced from MesADC (Dynamic cells) Effect from N R GS FDL P PBSt Activation Inhibition No effect
1 3 4
0
Number
8
3
0 2
0
0
0 0
3
2
5
DP 4 0
5
Extensor
Flexor 14
1
9
1
8 4
0 1
5
13
10
C . Cells not tested from MesADC
Effect from N R
FDL
P
ABSm Tib
Activation Inhibition
1
1
0
1 0
Number
1
1
0
1 0
1
1
PBSt
DP
Ten
Extensor
Flexor
3
1 0
4 0
8
0
4 1
3
5
1
4
9
1
MesADC-system only causes selective dynamic effects, all cells activated from this region are regarded as dynamic fusimotor neurones (23 cells in B). Of the other cells 27 were found to be influenced from NR but not from MesADC and these are regarded as static neurones (A). The third group in C consists of 13 cells which were tested only from NR and thus could not be characterized. A study of Table 11 reveals the following interesting points: 1 . the majority of dynamic cells were influenced also from NR. This influence was almost strictly reciprocal with excitation of flexor cells and inhibition of extensor cells (B). 2. a few dynamic cells were not reached by the rubrospinal tract (B). 3. among static cells to extensor muscles a mixture of excitation and inhibitory effects was found while flexor cells were always excited (A). 4. also within the material of unclassified cells (unknown MesADC-influence) excitatory effects on the flexor neurones predominate. Of four extensor cells, on the other hand, all were excited leading to the suggestion that at least these cells were all static, as excitatory effects on extensor dynamic neurones were rare (C). 11. The synaptic linkage between rubrospinal fibres and y-motoneurones
The technique of using long, continuous trains of stimuli in the central areas does not allow a determination of the latency of the effects evoked. Attempts to use shorter trains of stimuli were often made. From NR it was frequently noted, that cells activated from this region were influenced by trains containing 4 or 5 pulses. This influence, however, had the nature of irregular firing with greatly fluctuating latency from the effective shock and any exact appreciation of the segmental latency was impossible. In one experiment recording was made, not from cell bodies in the lumbar cord, but
436
B . APPELBERG. T. JENESKOG AND H. JOHANSSON
n
Th 13
IIIII I
Fig. 3. A, identification of tenuissimus fusimotor neurone recorded in nerve filament. In B and C three and four shocks in N R activate a as well as y motoneurone (uuuer . .. traces). Descending volley recorded in Th,, shown in lower traces. For further details see text. Time calibration 2 ms for A, 10 ms for B and 4 ms for C.
instead from axons in branches of the tenuissimus nerve. In this experiment a nerve filament was obtained containing a y-fibre with a conduction velocity of 47 m/s (Fig. 3 A). The fibre was spontaneously active (cf. arrows in A). The filament also contained some a-axons-two of these were activated together with the y-fibre from the L7-ventral root (Fig. 3 A), a third belonged to another ventral root. The y-fibre as well as the unidentified a-fibre were both activated from NR by a short train of stimuli. From the analysis of a number of superimposed records of the type illustrated in Fig. 3 B and C it became clear that the number of pulses required in the NR to activate the y-axon varied from record to record. Careful measurements of intervals between different stimulating pulses and the resulting burst of spikes did, however, reveal that the minimal latency from active pulse to fired y-spike was 7.8 ms. (In seven records each consisting of 10 or 20 superimposed sweeps thislatency to the burst was found from the second pulse in four cases and from the first, the third or the fourth pulse in one case each). In Fig. 3 B and C the cell is in both cases fired by the second rubrdi pulse. In this experiment activity in the fastest rubrospinal fibres reached the Th13segment in 2.0 ms (cf. arrows in Fig. 3 C, lower trace) and the L7-segment in 2.5 ms (additional distance 60 mm). Thus 5.3 ms remains for segmental synaptic delay and central and peripheral conduction. Conduction through the ventral root and the nerve could be calculated on the basis of the known conduction velocity of the fibre and the measured distance and was 3.5 ms. This leaves a segmental delay of 1.8 ms for the activation of this y-unit. This seems to imply a disynaptic linkage from the rubrospinal tract to this unit (cf. Discussion). More conclusive evidence concerning the synaptic linkage between the rubrospinal tract and a- as well as y-motoneurones was obtained in recordings from cells belonging to the DP-nerve in another experiment (Fig. 4). In both cells, the m conducting at 94.5 m/s and the y (classified as static and probably being recorded from in a juxtacellular position) at 14m/s, occasional EPSP:s could be evoked by two NR-pulses (B and G). Usually, however, three pulses were required to regularly depolarize the cells (C and H). When four or five stimuli were used these caused additional depolarization by their EPSP:s summating to the one caused by the third pulse (D and J). At still longer trains the y-cell started to fire as illustrated by the superimposed record in E. The latency from the active stimulating pulse to the resulting EPSP was for the a-cell 4.3 ms, for the y-cell 5.0 ms. The descending rubrospinal
NUCLEUS RUBER ON FUSIMOTOR NEURONES
437
I
C
H
4'
Fig. 4. Juxtacellular recording from a fusimotor (A-E) and intracellular recording from a skeletomotor (F-I) neurone both belonging to the deep peroneal nerve. In A and F the cells are antidromically identified. In B-D and G-l postsynaptic potentials caused by a varying number of stimuli in N R are shown. The short horizontal line below each record denotes the distance between the active stimulus and the resulting EPSP. In E the fusimotor neurone I S seen to start firing (superimposed records) to an increased number of rubral stimuli. In K is shown the descending rubrospinal volley in L,. Voltage calibration is 4 mV for record A, 0.4 mV for B-D, 10 mV for F-I. Time: 10 ms for B-E, 4 ms for A and F-K.
volley in this experiment reached the L7 segment in 3.1 ms (Fig. 4 K shows the volley recorded in L5 with a latency of 2.9 ms. About 0.2 ms should be added to account for the additional distance of 15 mm to the L7 segment). This leaves about 1.2 and 1.9 ms synaptic delay for the a- and the y-cell respectively. Also this observation indicates that a disynaptic coupling between the rubrospinal tract and a- as well as static y-cells may occur (cf. Discussion). During the course of the experiments a considerable number of intracellular recordings from a-motoneurones were obtained. On each such occasion the effect on the cell from the NR-stimulating electrode as well as from the MesADC was tested. It was consistently noted that while the rubrospinal activation regularly caused postsynaptic potentials in these cells according to the pattern described by Hongo et al. (1969 a) a-motoneurones were never influenced in any respect from the MesADC (cf. Discussion). 111. Conduction velocities of static and dynamic axons Within the whole material of 63 cells the conduction velocities of their axons varied from 14 to 55 m/s. For the 23 cells classified as dynamic the axon conduction velocity range was 21-47 m/s with a mean of 31.1 m/s, and for 26 static cells the corresponding figures were 17-35 m/s and 27.5 m/s. The tendency for the lowest conduction velocities to be found in static neurones and the highest in dynamic ones is clear but not statistically significant. It is, however, partly in agreement with the observations made by Brown, Crowe and Matthews (1965) who for the tibialis anterior muscle found that the slowest y-fibres (in their study with conduction velocities below 30 m/s) were virtually certain to be static. For the soleus muscle, on the other hand, Crowe and Matthews (1964) found no such clear differences in conduction
438
B. APPELBERG, T. JENESKOG A N D H. JOHANSSON
velocities between static and dynamic y-axons. With such differences between individual muscles it is clear that the matter of conduction velocities should not be too much emphasized in a material obtained from several muscles.
Discussion The primary aim of the present investigation was to study the effects of the rubrospinal tract on y-motoneurones. According to Baldissera et al. (1972 a and b) segmental rubrospinal effects may be studied in isolation after a bilateral ablation of the sensory-motor cortices if the stimulating electrode is placed so as to activate interposito-rubral fibres just ventrally to the NR. These precautions were taken in the present experiments and it seems safe to conclude that the effects studied were truly rubrospinal. Within the whole material of 30 extensor cells 1 1 were excited, 15 inhibited and 4 not influenced from the NR. Of 33 flexor cells 31 were excited, 1 inhibited and 1 not influenced. In a corresponding material of intracellularly recorded a-motoneurones (Hongo et al. 1969 a, their Table I) 45 out of 57 flexor motoneurones were purely excited, none were purely inhibited while 12 received a mixture of excitation and inhibition. Of 125 extensor motoneurones 50 were excited, 37 inhibited and 38 influenced in a mixed way. Although our material is smaller and obtained by extracellular recording it is immediately evident that there is a close correspondence in the pattern of rubrospinal control of a- and y-motoneurones. Having in mind the concept of a-y-linkage it is, however, clearly not enough to demonstrate a similarity in effect on the two groups of motoneurones from a certain source. It is also necessary to demonstrate an equality in minimal synaptic linkage. That the rubrospinal tract fibres may contact a-motoneurones via a single interneurone seems well established (Hongo et al. 1969 a and verified in the present investigation). In the present work on y-motoneurones, on the other hand, careful determination of time for central delay was arrived at only twice. For the juxtacellulary recorded DP-cell this delay was 1.9 rns, a figure possibly allowing a passage through three synapses. It is possible, however, to suggest a disynaptic linkage and ascribe the seemingly long delay to the possibility that the descending fibre activating the cell had a somewhat slower conduction velocity than the fastest ones causing the very beginning of the descending volley. For the y-fibre recorded in the tenuissimus muscle nerve the central delay could be calculated with satisfactory accuracy to be 1.8 ms. In this case, however, the time arrived at includes spike initiation and the disynaptic nature of the coupling can hardly be doubted. The fact that the cell in one recording actually responded to the first shock in NR may be attributed to the fact that the cell was spontaneously active and therefore sometimes extremely easily triggered. I n conclusion it is, therefore, provisionally suggested that the synaptic coupling between the rubrospinal tract and a- as well as static y-motoneurones is minimally disynaptic. Thereby the rubrospinal tract is added to the list of descending systems theoretically capable of providing a-y-linked movement. The vestibulo-spinal tract as well as the MLF-system were previously shown to be organized in this manner (I$. Grillner 1969). The second aim of the experiments was to attempt to divide the material of y-motoneurones
NUCLEUS RUBER ON FUSIMOTOR NEURONES
439
into the two functional classes, static and dynamic cells. The method adopted for this was based upon previous findings that the MesADC-region, i.e. the area just dorsally to and extending into the dorsal part of the red nucleus, selectively influences the dynamic sensitivity of muscle spindles (cf. Appelberg 1967, Appelberg and Jeneskog 1972 and also Appelberg and ErnonCt-Denand 1965 who showed that effects on extensor as well as flexor spindles were obtained from the same area). The work of Appelberg and Jeneskog (1972) is of special relevance in this connection. During their experiments various types of influence on single spindles could be observed from different levels in a stimulating electrode track passing through the mesencephalon at a rubral level. It became quite clear that the effects evoked from the area just dorsally to the red nucleus were always purely dynamic (in contrast to effects evoked from sites within the red nucleus, which were often mixed static/dynamic and, in the light of the present experiments, probably at times evoked via the rubrospinal tract). Jeneskog (1974 a) later made a detailed comparison between the mesencephalic area causing dynamic spindle effects and the one evoking a D-zone response in the cerebellar paramedian lobule and found a near to perfect coincidence between the two. In the present experiments the MesADC-electrode was placed with such a PM-response as a quide. At the same time it was carefully controlled that no rubrospinal descending volley was evoked from the chosen electrode site. With all this in mind we feel it fully justified to use such a MesADC-stimulus as a methodological tool enabling the separation of y-motoneurones into their two functional classes: static and dynamic. Accepting the method of classification leads to an important conclusion, namely that most dynamic y-motoneurones are also controlled by the rubrospinal tract. From this follows that the rubrospinal system may be thought of as a motor system working with linkage between a- and possibly both classes of y-motoneurones. Nothing is known, however, about the synaptic coupling between the tract and dynamic neurones. It may be of special relevance also to note that of 23 cells classified as dynamic 5 were not influenced from the NR (Table I1 B). This may indicate that there exists a population of dynamic ymotoneurones not participating in a- y-linked control but rather being private to the specialized dynamic control system. Whether other such subpopulations of dynamic y-motoneurones exist possibly being private to other descending or reflex systems is at present not known. Appelberg and Jeneskog (1972) did, however, note that not all spindles seemed to have their dynamic sensitivity controlled by the MesADC system, It should finally be pointed out, that the observations of lack of any effect on a-motoneurones from the MesADC (while synaptic potentials were readily evoked from the NR) are of considerable importance. They do strengthen the notion that the MesADC-system, which according to Jeneskog (1974 a, b) may be a rubro-bulbo-spinal system with its origin in the rostra1 parts of the N R , is a system exclusively controlling one type of rnotoneurones, the dynamic fusimotor ones. Whether this system may also have other neuronal connections in the spinal cord is not known. In fact a descending volley reminding of the one caused by the dorsal reticulo-spinal system (Baldissera et al. 1972 a) was sometimes observed when stimulating in the MesADC. The possible identity of the MesADC-system and the dorsal reticulo-spinal system remains an object for further investigation. This work was supported by the Swedish Medical Research Council, Project No. 3873.
440
B. APPELBERG, T. JENESKOG A N D H. JOHANSSON
References APPELBEKC, B., The effect of electrical stimulation in nucleus ruber on the response to stretch in primary and secondary muscle spindle afferents. Acra physiol. scancl. 1962. 56. 140-151. APPELBEKC, B., A rubro-olivary pathway. 11. Simultaneous action on dynamic fusimotor neurones and the activity of the posterior lobe of the cerebellar cortex. Exp. Brain Res. 1967. 3. 382-390. APPELBERG, B. and F. EMON~T-DENAND, Central control of static and dynamic sensitivities of muscle spindle primary endings. Acra phjuiol. scand. 1965. 63. 487-494. APPELREKG, B. and I. Z . KOSARY, Excitation of flexor fusimotor neurones by electrical stimulation in the red nucleus. Artu physiol. scand. 1963. 59. 445-453. APPELBERC, B. and T. JENEsKoc, Mesencephalic fusimotor control. Exp. Brain Res. 1972. 15. 97-1 12. BALDISSERA. F., A. LUNDBERC and M. Uoo, Activity evoked from the mesencephalic tegmentum in descending pathways other than the rubrospinal tract. Exp. Brain Res. 1972 a. 15. 133-150. BALDISSERA, F., A. LUNDBERG and M. UDO, Stimulation of pre- and postsynaptic elements in the red nucleus. Exp. Bruin Rus. 1972 b. 15. 151-167. BROWN,M. C., A. CROWEand P. B. C. MATTHEWS, Observations on the fusimotor fibres of the tibialis poslerior muscle of the cat. J . Physiol. (Lond.) 1965. 177. 140-159. CROWE, A . and P. B. C. MATTHEWS, Further studies of static and dynamic fusimotor fibres. J. Physiol. (Lond.) 1964. 174. 132-151. GRILLNER, S . , Supraspinal and segmental control of static and dynamic y-motoneurones in the cat. Acta physiol. scand. 1969. Suppl. 327. HONGO,T., E L ~ B I E TJANKOWSKA A and A. LUNDBERG, The rubrospinal tract. 1. Effects on alpha-motoneurones innervating hindlimb muscles of the cat. Exp. Brain Res. 1969 a. 7. 344-364. HONW,T., ELZBIETA JANKOWSKAand A. LUNDBEKG, The rubrospinal tract. 11. Facilitation of interneuronal transmission in reflex paths to motoneurones. Exp. Brain Res. 1969 b. 7. 365-391. HONGO,T., ELZBIETA JANKOWSKAand A. LUNDBERC, The rubrospinal tract. 111. Effects on primary afferent terminals. Exp. Bruin Res. 1972 a. 15. 39-53. HONGO,T., ELZRIETAJANKOWSKA and A. LUNDBEKC, The rubrospinal tract. I V . Effects on interneurones. Exp. Brain Res. 1972 b. 15. 54-78. JENESKOC, T., Parallel activation of dynamic fusimotor neurones and a climbing fibre system from the cat brain stem. 1. Effects from the rubral region. Acta physiol. scand. 1974 a. 91. 223-242. JFNESKOC. T., Parallel activation of dynamic fusimotor neurones and a climbing fibre system from the cat brain stem. I I . Effects from the inferior olivary region. Actaphysiol. scund. 1974 h. 92. 66-83. SASAKI, K., A. NAMIKAWA and S. HASHIRAMOTO, The effect of midbrain stimulation upon alpha-niotoneurones i n lumhar spinal cord of the cat. Jap. J . Physiol. 1960. 10. 303-316.
Rubrospinal Control of Static and Dynamic Fusimotor Neurones BY
B. APPELBERG, T. JENESKOG and H. JOHANSSON Received 7 May 1975
Abstract APPELBERG, B., T. JENESKOG and H. JOHANSSON. Rubrospinal control of static and dynamic fusimotor neurones. Acta physiol. scand. 1975. 95. 431-440. Rubrospinal effects on about 60 extracellularly recorded y-motoneurones were studied in anesthetized cats. All cells were antidromically identified from various muscle nerves. 23 cells were regarded as dynamic as they were activated from a mesencephalic region previously known to influence selectively muscle spindle dynamic sensitivity. The pattern of rubrospinal influence on static fusimotor neurones to different muscles closely followed that previously demonstrated for a-motoneurones with predominantly excitation of flexor neurones and excitation or inhibition in equal amounts of extensor cells. Dynamic fusimotor neurones were influenced in a strictly reciprocal manner with excitation of flexor cells and inhibition of extensor cells except for a few neurones which could not be reached from nucleus ruber. Evidence was also obtained indicating that the shortest path from nucleus ruber to static fusimotor neurones involves one interneurone.
The effect of electrical stimulation within the red nucleus on the fusimotor system was studied by Appelberg (1962) and by Appelberg and Kosary (1963) by the indirect method of recording from muscle spindle afferents. It was regularly observed that such stimulation caused inhibition of spontaneous activity in extensor spindles while flexor spindles were instead excited. These findings were in general agreement with the notion at that time that the red nucleus was excitatory to flexor and inhibitory to extensor muscles (Sasaki, Namikawa and Hashiramoto 1960). In recent years the control function of the rubrospinal tract as well as of other systems descending from the mesencephalon and acting on the neuronal machinery in lumbar segments have been extensively and carefully studied by Lundberg and coworkers (cf. Hongo, Jankowska and Lundberg 1969 a, b; 1972 a, b and Baldissera, Lundberg and Udo 1972 a, b). These studies have, among other things, given extremely detailed information about rubrospinal control of hind limb a-motoneurones and also carefully analyzed methodological aspects of importance when using electrical stimulation in the rubral region. With this in mind and considering current theories of linkage between a- and y-motoneurones in motor control it seemed of great interest to reinvestigate in greater detail and 43 1
432
B. APPELBERG, T. JENESKOG AND H. JOHANSSON
with another technique the matter of rubrospinal control of hind limb fusimotor neurones. A detailed knowledge of a descending system selectively influencing dynamic fusimotor neurones (cf. Appelberg and Jeneskog 1972, Jeneskog 1974 a) was utilized as a methodological tool allowing the separation of y-motoneurones into the two functional groups, static and dynamic.
Methods The results to be presented were obtained in experiments on 18 cats. Of these 11 were anesthetized with halothane (usually 1.0-1.5%) administered in a mixture of Oxygen (1/3) and Nitrous Oxide (2/3), 4 with a-Chloralose (60 mg/kg) administered intravenously after induction with halothane and 3 with pentobarbital (40 mg/kg) given intraperitoneally with later supplementary intravenous doses. Essentially similar findings were obtained regardless of type of anesthesia. Blood pressure and expiratory CO, as well as rectal temperature and temperatures in the paraffin pools were continuously monitored throughout the experiments.
Operation The operative procedure included a low thoracic and a lumbar laminectomy. Th e dorsal columns and the right spinal half were sectioned in Th12-13. The lumbar spinal cord was prepared for microelectrode recording from himd limb motoneurone pools. In the left hind limb the following nerves were dissected and mounted for stimulation: posterior bicepssemitendinosus (PBSt), anterior biceps-semimembranosus (ABSm), gastrocnemius-soleus (GS),plantaris (P),flexor digitorum and hallucis longus (FDL), deep peroneal (DP), superficial peroneal (SP) and tibia1 (Tib). In one experiment the tenuissimus nerve was prepared instead and later used for dissection of filaments (see Results, Fig. 3). The sensorimotor cortex was acutely ablated bilaterally in all experiments. Access to the right mesencephalon for stimulating electrodes and to the left paramedian lobule of the cerebellum for surface recording was arranged.
Recording Extracellular recording was made from y-motoneurones with glass capillary electrodes filled with 4 M NaCI. The impedance of the electrodes was usually about 1-2 M a . Only cells which could be antidromically invaded from one of the dissected muscle nerves and thus identified were accepted. The conduction velocities of the axons could then be determined by also measuring the conduction distance along the different nerves after each experiment. Only cells having a spontaneous activity could be tested with regard to inhibitory influence from the brain, and therefore each cell included in the material exhibited such an activity during at least part of the recording session. Surface recording of the rubrospinal tract volley on the dorsolaterd aspect of the cord in the exposed thoracic segment was employed to guide the placement of a stimulating electrode in the red nucleus (NR). Surface recording of climbing fibre responses in the D-zone of the cerebellar paramedian lobule (PM) was employed when placing a stimulating electrode in a mesencephalic area previously shown to selectively influence dynamic fusimotor neurones (MesADC-cf. Appelberg 1967 and Jeneskog 1974 a and Discussion).
Stimulation For stimulation in the brain two medio-lateral grids of 3 glasscovered platinum-iridium electrodes with 1 mm interelectrode distance were used. One of the grids was angled 15" to the vertical plane as all electrodes had to be positioned at approximately the same rostro-caudal level but at different depths. A presumably selective activation of the rubro-spinal tract was obtained by stimulation of interpositorubral fibres in the caudo-ventral part of or just ventrally to the caudal part of the red nucleus. The correct placement of the stimulating electrode was aided by observing the descending volley change from a short latency direct response to a transsynaptic response with longer latency when slowly proceeding with the electrode through the red nucleus (cf. Baldissera et al. 1972 b). This placement allowed a maximal activation of the rubro-spinal tract at low stimulating strength at a considerable distance away from the MesADC.
433
NUCLEUS RUBER ON FUSIMOTOR NEURONES
P-Y 31 rn/s
NR90pA
C
B
-
E F
NR 3 0 p A
I
I
MesADC 6 0 p A
o.5s
Fig. 2
Fig. I
Fig. 1. A and B are camera lucida drawings of transverse sections through the mesencephalon at a rubral level (B about I mm caudal to A) with stimulating sites indicated. Three stimulating shocks at 600 Hz at the site in A activates rubrospinal tract as revealed by surface recording from the cord in C (upper trace). The stimulating point in B causes no rubrospinal volley (D, upper trace) but instead a climbing fibre response in the paramedian lobule of the cerebellum is evoked (D, lower trace). Positivity is recorded upwards in this and all subsequent figures. Time calibration 5 ms, voltage calibration 100 pV to upper traces in C and D, 200 ,uV to lower traces. Fig. 2. A, B, identification of plantaris fusimotor neurone (see text). C , descending volley in Th,, to N R stimulation. D, PM-response evoked from MesADC (upper trace). Note lack of descending volley (lower trace). Time calibration below B is 10 ms to A, B and D, 4 ms to C . Voltage calibration is IOOyV for upper trace in D. E, inhibition of spontaneous fusimotor activity by repetitive stimulation in N R during time indicated by black line below record. F, activation of same fusimotor neurone from MesADC.
An equally selective activation of the descending pathway to dynamic fusimotor neurones was attempted by placing a second stimulating electrode dorsally in the low-threshold region for a D-zone climbing fibre response in the PM. A direct activation of rubrospinal cells could usually be obtained from this region, but only at high stimulating intensities. Only effects on y-motoneurones at strengths well below the threshold for NR-activation were considered. Also at such lower stimulating strengths a descending volley could sometimes be recorded in the thoracic segment during train stimulation. This volley was similar to that described by Baldissera el al. (1972 a) as being characteristic for the dorsal reticulo-spinal system. In Fig. I the position of stimulating sites for rubro-spinal activation (A) and within the MesADC (B) in one experiment, where histological control was made, is illustrated. In C and D the physiological criteria used for correct placement in all experiments are illustrated. In a few early experiments only NR-stimulation took place and of the total cell-material of 63 cells 13 were thus not tested from MesADC (see under Results).
Results 1. Effects from N R and MesADC on y-motoneurones
Table I summarizes the results obtained in one experiment (same experiment as Fig. 1). Nine y-motoneurones were recorded from, 6 of which were completely tested with regard to 33 - 755882
434
B. APPELBERG, T. JENESKOG AND H. JOHANSSON
TABLE 1. Effects from nucleus ruber (NR) and mesencephalic area for dynamic control of muscle spindles (MesADC) on nine fusirnotor neurones in 1 expt. Thresholds for effects given in PA. t actrvation, - inhibition, O=no influence. ~
~
Cell no. ~~
~~
Type
1
FDL
DP DP DP DP DP DP PBSt FDL
6
NR
PA
-
50 50 60 30 40 40 70 60
MesADC pA
~~
2 3 4 5 7 8 9
-
~~~
Cond. vel. 32 m/s 32 rn/s 21 m/s 34 m/s 21 m/s 30 rn/s 23 m/s 16mjs
+ +
+ i+ -t+ -
+ + + + +
0
60 70 70 60 70 40
0
Threshold for rubrospinal volley from NR 30 PA. Threshold for rubrospinal volley from MesADC 90 / L A . Threshold for PM-response from MesADC 30 /LA.
effects evoked from the two central regions, while 3 cells were kept only for short periods of time and not fully tested. The threshold for evoking a rubrospinal volley from the electrode just ventrally to the N R was about 30 PA. From the electrode placed in MesADC a weak rubrospinal volley could be evoked at stimulating strengths above 90 PA. A PM-response was, on the other hand, evoked already at 30 pA. It is clear from the Table that all effects on the different cells were obtained at stimulating intensities which makes spread of current from one region to the other unlikely. This conclusion is supported by the fact that cell nr I , which was reciprocally influenced from the two regions at low stimulating intensities, was controlled in the same way, but stronger, when stimulation was increased to 90 ,uA in N R and to 80 pA in MesADC. The central effects on all cells included in the material presented have been studied under similar circumstances. A cell influenced in the same reciprocal way as the FDL-y (cell 1) in Table I but belonging to the plantaris muscle is illustrated in Fig. 2. This cell was slightly difficult to identify as it responded antidromically from the nerve together with another cell (A). From the recordings in A and B it appears, however, that when a spontaneous spike of the cell occurred closely enough to the antidromic stimulus (B) a collision of orthodromic and antidromic spikes took place in the axon and the antidromic response disappeared. The remaining smaller cell belonging to the same nerve was not spontaneously active, nor was it influenced by the central stimuli. The rubrospinal volley evoked from the NR-electrode is illustrated in C , and the characteristic PM-response from the MesADC-electrode in D (note lack of volley in Th 13). In Fig. 2 E is seen a strong inhibitory influence on the spontaneously active cell from NR at low stimulus intensity. In F the cell is instead shown to be strongly activated from MesADC. In this experiment the thresholds for detectable descending rubrospinal volley from the NR-electrode and for PM-response from the MesADC-electrode were both around 40 [LA.A weak rubrospinal volley for MesADC-stimulation was seen at 100 PA. In Table 11 the whole material of 63 cells and the effects from MesADC and NR is presented. The cells have been divided into 3 subpopulations on the basis of the effects upon them from MesADC. As previous work on muscle spindle afferents has revealed that the
435
NUCLEUS RUBER ON FUSIMOTOR NEURONES
TABLE 11. Effects from NR and MesADC on whole material of 63 cells. A. Cells not influenced from MesADC (Static cells) Effect from N R GS FDL P ABSm
PBSt
DP
Activation Inhibition
5
6
0
I
0
5
13
14
Extensor
Flexor
Number
0
1
5
0 1
1
0
9 0
10
1
1
I
9
5
B. Cells influenced from MesADC (Dynamic cells) Effect from N R GS FDL P PBSt Activation Inhibition No effect
1 3 4
0
Number
8
3
0 2
0
0
0 0
3
2
5
DP 4 0
5
Extensor
Flexor 14
1
9
1
8 4
0 1
5
13
10
C . Cells not tested from MesADC
Effect from N R
FDL
P
ABSm Tib
Activation Inhibition
1
1
0
1 0
Number
1
1
0
1 0
1
1
PBSt
DP
Ten
Extensor
Flexor
3
1 0
4 0
8
0
4 1
3
5
1
4
9
1
MesADC-system only causes selective dynamic effects, all cells activated from this region are regarded as dynamic fusimotor neurones (23 cells in B). Of the other cells 27 were found to be influenced from NR but not from MesADC and these are regarded as static neurones (A). The third group in C consists of 13 cells which were tested only from NR and thus could not be characterized. A study of Table 11 reveals the following interesting points: 1 . the majority of dynamic cells were influenced also from NR. This influence was almost strictly reciprocal with excitation of flexor cells and inhibition of extensor cells (B). 2. a few dynamic cells were not reached by the rubrospinal tract (B). 3. among static cells to extensor muscles a mixture of excitation and inhibitory effects was found while flexor cells were always excited (A). 4. also within the material of unclassified cells (unknown MesADC-influence) excitatory effects on the flexor neurones predominate. Of four extensor cells, on the other hand, all were excited leading to the suggestion that at least these cells were all static, as excitatory effects on extensor dynamic neurones were rare (C). 11. The synaptic linkage between rubrospinal fibres and y-motoneurones
The technique of using long, continuous trains of stimuli in the central areas does not allow a determination of the latency of the effects evoked. Attempts to use shorter trains of stimuli were often made. From NR it was frequently noted, that cells activated from this region were influenced by trains containing 4 or 5 pulses. This influence, however, had the nature of irregular firing with greatly fluctuating latency from the effective shock and any exact appreciation of the segmental latency was impossible. In one experiment recording was made, not from cell bodies in the lumbar cord, but
436
B . APPELBERG. T. JENESKOG AND H. JOHANSSON
n
Th 13
IIIII I
Fig. 3. A, identification of tenuissimus fusimotor neurone recorded in nerve filament. In B and C three and four shocks in N R activate a as well as y motoneurone (uuuer . .. traces). Descending volley recorded in Th,, shown in lower traces. For further details see text. Time calibration 2 ms for A, 10 ms for B and 4 ms for C.
instead from axons in branches of the tenuissimus nerve. In this experiment a nerve filament was obtained containing a y-fibre with a conduction velocity of 47 m/s (Fig. 3 A). The fibre was spontaneously active (cf. arrows in A). The filament also contained some a-axons-two of these were activated together with the y-fibre from the L7-ventral root (Fig. 3 A), a third belonged to another ventral root. The y-fibre as well as the unidentified a-fibre were both activated from NR by a short train of stimuli. From the analysis of a number of superimposed records of the type illustrated in Fig. 3 B and C it became clear that the number of pulses required in the NR to activate the y-axon varied from record to record. Careful measurements of intervals between different stimulating pulses and the resulting burst of spikes did, however, reveal that the minimal latency from active pulse to fired y-spike was 7.8 ms. (In seven records each consisting of 10 or 20 superimposed sweeps thislatency to the burst was found from the second pulse in four cases and from the first, the third or the fourth pulse in one case each). In Fig. 3 B and C the cell is in both cases fired by the second rubrdi pulse. In this experiment activity in the fastest rubrospinal fibres reached the Th13segment in 2.0 ms (cf. arrows in Fig. 3 C, lower trace) and the L7-segment in 2.5 ms (additional distance 60 mm). Thus 5.3 ms remains for segmental synaptic delay and central and peripheral conduction. Conduction through the ventral root and the nerve could be calculated on the basis of the known conduction velocity of the fibre and the measured distance and was 3.5 ms. This leaves a segmental delay of 1.8 ms for the activation of this y-unit. This seems to imply a disynaptic linkage from the rubrospinal tract to this unit (cf. Discussion). More conclusive evidence concerning the synaptic linkage between the rubrospinal tract and a- as well as y-motoneurones was obtained in recordings from cells belonging to the DP-nerve in another experiment (Fig. 4). In both cells, the m conducting at 94.5 m/s and the y (classified as static and probably being recorded from in a juxtacellular position) at 14m/s, occasional EPSP:s could be evoked by two NR-pulses (B and G). Usually, however, three pulses were required to regularly depolarize the cells (C and H). When four or five stimuli were used these caused additional depolarization by their EPSP:s summating to the one caused by the third pulse (D and J). At still longer trains the y-cell started to fire as illustrated by the superimposed record in E. The latency from the active stimulating pulse to the resulting EPSP was for the a-cell 4.3 ms, for the y-cell 5.0 ms. The descending rubrospinal
NUCLEUS RUBER ON FUSIMOTOR NEURONES
437
I
C
H
4'
Fig. 4. Juxtacellular recording from a fusimotor (A-E) and intracellular recording from a skeletomotor (F-I) neurone both belonging to the deep peroneal nerve. In A and F the cells are antidromically identified. In B-D and G-l postsynaptic potentials caused by a varying number of stimuli in N R are shown. The short horizontal line below each record denotes the distance between the active stimulus and the resulting EPSP. In E the fusimotor neurone I S seen to start firing (superimposed records) to an increased number of rubral stimuli. In K is shown the descending rubrospinal volley in L,. Voltage calibration is 4 mV for record A, 0.4 mV for B-D, 10 mV for F-I. Time: 10 ms for B-E, 4 ms for A and F-K.
volley in this experiment reached the L7 segment in 3.1 ms (Fig. 4 K shows the volley recorded in L5 with a latency of 2.9 ms. About 0.2 ms should be added to account for the additional distance of 15 mm to the L7 segment). This leaves about 1.2 and 1.9 ms synaptic delay for the a- and the y-cell respectively. Also this observation indicates that a disynaptic coupling between the rubrospinal tract and a- as well as static y-cells may occur (cf. Discussion). During the course of the experiments a considerable number of intracellular recordings from a-motoneurones were obtained. On each such occasion the effect on the cell from the NR-stimulating electrode as well as from the MesADC was tested. It was consistently noted that while the rubrospinal activation regularly caused postsynaptic potentials in these cells according to the pattern described by Hongo et al. (1969 a) a-motoneurones were never influenced in any respect from the MesADC (cf. Discussion). 111. Conduction velocities of static and dynamic axons Within the whole material of 63 cells the conduction velocities of their axons varied from 14 to 55 m/s. For the 23 cells classified as dynamic the axon conduction velocity range was 21-47 m/s with a mean of 31.1 m/s, and for 26 static cells the corresponding figures were 17-35 m/s and 27.5 m/s. The tendency for the lowest conduction velocities to be found in static neurones and the highest in dynamic ones is clear but not statistically significant. It is, however, partly in agreement with the observations made by Brown, Crowe and Matthews (1965) who for the tibialis anterior muscle found that the slowest y-fibres (in their study with conduction velocities below 30 m/s) were virtually certain to be static. For the soleus muscle, on the other hand, Crowe and Matthews (1964) found no such clear differences in conduction
438
B. APPELBERG, T. JENESKOG A N D H. JOHANSSON
velocities between static and dynamic y-axons. With such differences between individual muscles it is clear that the matter of conduction velocities should not be too much emphasized in a material obtained from several muscles.
Discussion The primary aim of the present investigation was to study the effects of the rubrospinal tract on y-motoneurones. According to Baldissera et al. (1972 a and b) segmental rubrospinal effects may be studied in isolation after a bilateral ablation of the sensory-motor cortices if the stimulating electrode is placed so as to activate interposito-rubral fibres just ventrally to the NR. These precautions were taken in the present experiments and it seems safe to conclude that the effects studied were truly rubrospinal. Within the whole material of 30 extensor cells 1 1 were excited, 15 inhibited and 4 not influenced from the NR. Of 33 flexor cells 31 were excited, 1 inhibited and 1 not influenced. In a corresponding material of intracellularly recorded a-motoneurones (Hongo et al. 1969 a, their Table I) 45 out of 57 flexor motoneurones were purely excited, none were purely inhibited while 12 received a mixture of excitation and inhibition. Of 125 extensor motoneurones 50 were excited, 37 inhibited and 38 influenced in a mixed way. Although our material is smaller and obtained by extracellular recording it is immediately evident that there is a close correspondence in the pattern of rubrospinal control of a- and y-motoneurones. Having in mind the concept of a-y-linkage it is, however, clearly not enough to demonstrate a similarity in effect on the two groups of motoneurones from a certain source. It is also necessary to demonstrate an equality in minimal synaptic linkage. That the rubrospinal tract fibres may contact a-motoneurones via a single interneurone seems well established (Hongo et al. 1969 a and verified in the present investigation). In the present work on y-motoneurones, on the other hand, careful determination of time for central delay was arrived at only twice. For the juxtacellulary recorded DP-cell this delay was 1.9 rns, a figure possibly allowing a passage through three synapses. It is possible, however, to suggest a disynaptic linkage and ascribe the seemingly long delay to the possibility that the descending fibre activating the cell had a somewhat slower conduction velocity than the fastest ones causing the very beginning of the descending volley. For the y-fibre recorded in the tenuissimus muscle nerve the central delay could be calculated with satisfactory accuracy to be 1.8 ms. In this case, however, the time arrived at includes spike initiation and the disynaptic nature of the coupling can hardly be doubted. The fact that the cell in one recording actually responded to the first shock in NR may be attributed to the fact that the cell was spontaneously active and therefore sometimes extremely easily triggered. I n conclusion it is, therefore, provisionally suggested that the synaptic coupling between the rubrospinal tract and a- as well as static y-motoneurones is minimally disynaptic. Thereby the rubrospinal tract is added to the list of descending systems theoretically capable of providing a-y-linked movement. The vestibulo-spinal tract as well as the MLF-system were previously shown to be organized in this manner (I$. Grillner 1969). The second aim of the experiments was to attempt to divide the material of y-motoneurones
NUCLEUS RUBER ON FUSIMOTOR NEURONES
439
into the two functional classes, static and dynamic cells. The method adopted for this was based upon previous findings that the MesADC-region, i.e. the area just dorsally to and extending into the dorsal part of the red nucleus, selectively influences the dynamic sensitivity of muscle spindles (cf. Appelberg 1967, Appelberg and Jeneskog 1972 and also Appelberg and ErnonCt-Denand 1965 who showed that effects on extensor as well as flexor spindles were obtained from the same area). The work of Appelberg and Jeneskog (1972) is of special relevance in this connection. During their experiments various types of influence on single spindles could be observed from different levels in a stimulating electrode track passing through the mesencephalon at a rubral level. It became quite clear that the effects evoked from the area just dorsally to the red nucleus were always purely dynamic (in contrast to effects evoked from sites within the red nucleus, which were often mixed static/dynamic and, in the light of the present experiments, probably at times evoked via the rubrospinal tract). Jeneskog (1974 a) later made a detailed comparison between the mesencephalic area causing dynamic spindle effects and the one evoking a D-zone response in the cerebellar paramedian lobule and found a near to perfect coincidence between the two. In the present experiments the MesADC-electrode was placed with such a PM-response as a quide. At the same time it was carefully controlled that no rubrospinal descending volley was evoked from the chosen electrode site. With all this in mind we feel it fully justified to use such a MesADC-stimulus as a methodological tool enabling the separation of y-motoneurones into their two functional classes: static and dynamic. Accepting the method of classification leads to an important conclusion, namely that most dynamic y-motoneurones are also controlled by the rubrospinal tract. From this follows that the rubrospinal system may be thought of as a motor system working with linkage between a- and possibly both classes of y-motoneurones. Nothing is known, however, about the synaptic coupling between the tract and dynamic neurones. It may be of special relevance also to note that of 23 cells classified as dynamic 5 were not influenced from the NR (Table I1 B). This may indicate that there exists a population of dynamic ymotoneurones not participating in a- y-linked control but rather being private to the specialized dynamic control system. Whether other such subpopulations of dynamic y-motoneurones exist possibly being private to other descending or reflex systems is at present not known. Appelberg and Jeneskog (1972) did, however, note that not all spindles seemed to have their dynamic sensitivity controlled by the MesADC system, It should finally be pointed out, that the observations of lack of any effect on a-motoneurones from the MesADC (while synaptic potentials were readily evoked from the NR) are of considerable importance. They do strengthen the notion that the MesADC-system, which according to Jeneskog (1974 a, b) may be a rubro-bulbo-spinal system with its origin in the rostra1 parts of the N R , is a system exclusively controlling one type of rnotoneurones, the dynamic fusimotor ones. Whether this system may also have other neuronal connections in the spinal cord is not known. In fact a descending volley reminding of the one caused by the dorsal reticulo-spinal system (Baldissera et al. 1972 a) was sometimes observed when stimulating in the MesADC. The possible identity of the MesADC-system and the dorsal reticulo-spinal system remains an object for further investigation. This work was supported by the Swedish Medical Research Council, Project No. 3873.
440
B. APPELBERG, T. JENESKOG A N D H. JOHANSSON
References APPELBEKC, B., The effect of electrical stimulation in nucleus ruber on the response to stretch in primary and secondary muscle spindle afferents. Acra physiol. scancl. 1962. 56. 140-151. APPELBEKC, B., A rubro-olivary pathway. 11. Simultaneous action on dynamic fusimotor neurones and the activity of the posterior lobe of the cerebellar cortex. Exp. Brain Res. 1967. 3. 382-390. APPELBERG, B. and F. EMON~T-DENAND, Central control of static and dynamic sensitivities of muscle spindle primary endings. Acra phjuiol. scand. 1965. 63. 487-494. APPELREKG, B. and I. Z . KOSARY, Excitation of flexor fusimotor neurones by electrical stimulation in the red nucleus. Artu physiol. scand. 1963. 59. 445-453. APPELBERC, B. and T. JENEsKoc, Mesencephalic fusimotor control. Exp. Brain Res. 1972. 15. 97-1 12. BALDISSERA. F., A. LUNDBERC and M. Uoo, Activity evoked from the mesencephalic tegmentum in descending pathways other than the rubrospinal tract. Exp. Brain Res. 1972 a. 15. 133-150. BALDISSERA, F., A. LUNDBERG and M. UDO, Stimulation of pre- and postsynaptic elements in the red nucleus. Exp. Bruin Rus. 1972 b. 15. 151-167. BROWN,M. C., A. CROWEand P. B. C. MATTHEWS, Observations on the fusimotor fibres of the tibialis poslerior muscle of the cat. J . Physiol. (Lond.) 1965. 177. 140-159. CROWE, A . and P. B. C. MATTHEWS, Further studies of static and dynamic fusimotor fibres. J. Physiol. (Lond.) 1964. 174. 132-151. GRILLNER, S . , Supraspinal and segmental control of static and dynamic y-motoneurones in the cat. Acta physiol. scand. 1969. Suppl. 327. HONGO,T., E L ~ B I E TJANKOWSKA A and A. LUNDBERG, The rubrospinal tract. 1. Effects on alpha-motoneurones innervating hindlimb muscles of the cat. Exp. Brain Res. 1969 a. 7. 344-364. HONW,T., ELZBIETA JANKOWSKAand A. LUNDBEKG, The rubrospinal tract. 11. Facilitation of interneuronal transmission in reflex paths to motoneurones. Exp. Brain Res. 1969 b. 7. 365-391. HONGO,T., ELZBIETA JANKOWSKAand A. LUNDBERC, The rubrospinal tract. 111. Effects on primary afferent terminals. Exp. Bruin Res. 1972 a. 15. 39-53. HONGO,T., ELZRIETAJANKOWSKA and A. LUNDBEKC, The rubrospinal tract. I V . Effects on interneurones. Exp. Brain Res. 1972 b. 15. 54-78. JENESKOC, T., Parallel activation of dynamic fusimotor neurones and a climbing fibre system from the cat brain stem. 1. Effects from the rubral region. Acta physiol. scand. 1974 a. 91. 223-242. JFNESKOC. T., Parallel activation of dynamic fusimotor neurones and a climbing fibre system from the cat brain stem. I I . Effects from the inferior olivary region. Actaphysiol. scund. 1974 h. 92. 66-83. SASAKI, K., A. NAMIKAWA and S. HASHIRAMOTO, The effect of midbrain stimulation upon alpha-niotoneurones i n lumhar spinal cord of the cat. Jap. J . Physiol. 1960. 10. 303-316.
