- Email: [email protected]
Monosynaptic activation of long descending propriospinal neurons from the lateral vestibular nucleus and the medial longitudinal fasciculus
EXPERIMENTAL
NEUROLOGY
86.462-472
(1984)
Monosynaptic Activation of Long Descending Propriospinal Neurons from the Lateral Vestibular Nucleus and the Medial Longitudinal Fasciculus R. D. SKINNER, Department
R. S. REMMEL,
AND L. B. MINOR’
of Anatomy and Department of Physiology and Biophysics Arkansas for Medical Sciences. Little Rock, Arkansas Received
March
27, 1984: revision
received
(R.S.R.), 72205
University
of
July 27, I984
In decerebrate cats long descending propriospinal (LDP) neurons were recorded extracellularly in the cervical enlargement. They were identified antidromically by spinal cord stimulation at the Ll -L2 level. Inputs to these cells were tested by stimulating the medial longitudinal fasciculus (MLF) 5 to 6 mm rostra1 to the obex. the lateral vestibular nucleus (LVN). the upper MLF I mm caudal to the trochlear nucleus, and the medial vestibular nucleus (MVN), all on the ipsilateral side. Action potentials were elicited in 44% (64/144) of LDP neurons in the ventral horn (laminae VII, VIII) at a segmental latency of 1 ms or less following brain stem stimulation. This was considered to be a monosynaptic latency. The most effective stimulation sites were the MLF and the LVN. MLF stimulation accounted for about two-thirds of the monosynaptically elicited action potentials and LVN for about one-third. Another 22% of LDP neurons responded at longer latencies, but some of those responses may also have been monosynaptic. Stimulation of the upper MLF and the MVN were much less effective,indicating that the MLF input was predominantly from fibers originating in the medullary and/or pontine reticular formation. Q 1984 Academic
Press. Inc.
INTRODUCTION Propriospinal neurons connecting different segments of the spinal cord constitute a heterogeneous population of neurons having different lengths Abbreviations: LDP-long descending propriospinal: MLF-medial longitudinal fasciculus: LVN, MVN-lateral, medial vestibular nucleus; Nr-nucleus reticularis; VC-ventral cord; PS-propriospinal. ’ We thank Nancy Stone for typing and Gael Sammartino for technical assistance. David Borne participated in some of the experiments. Lloyd B. Minor was an extem from Brown University, Division of Biology and Medicine. His present address is Department of Surgery, Duke University School of Medicine. Durham, NC 27705. This work was sponsored by U.S. Public Health Service grant NS 10304 and by National Science Foundation grant ISP 801 1447. 462 OOl4-4886184 $3.00 Copyright 0 1984 by Academic Press. Inc. All n&s of rcproductron I” any form rescrred.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
463
of axons and, most likely, a variety of functions. The lengths of their axons vary from one segment to possibly the length of the spinal cord (19, 23, 30, 36). They function in spinal reflexes in cats (15-l 7, 22, 27), in dogs (28), and probably in man (5, 21). An intriguing, but untested, hypothesis of the function of long propriospinal neurons is that they are part of the spinal locomotion generators and help to coordinate the rostra1 and caudal pairs of limbs during locomotion in tetrapods (7, 22). The purpose of this study was to extend previous observations regarding descending input to long descending propriospinal (LDP) neurons. The LDP neurons studied had their cell bodies in the cervical enlargement and projected to the lumbosacral region (30). Activation at short latencies, particularly those short enough for a monosynaptic connection, would indicate that these cells mediate directly the effects of some descending pathways. A brief description of this work has been reported (32). METHODS In 25 cats under ether anesthesia, the carotid arteries were ligated and a precollicular decerebration was done. Blood pressure was maintained above 90 mm Hg by means of an infusion of norepinephrine (80 pg/ml). Rectal temperature and end-tidal Pcoz were monitored and maintained near normal values. The occipital bone and midline cerebellum were removed to expose the fourth ventricle. Laminectomies were made at C5-Tl and Ll -L2. The cat’s head was held in a stereotaxic frame and the T2 spinous process was rigidly clamped. The cat was paralyzed with gallamine triethiodide and artificially ventilated. A bilateral pneumothorax reduced movements. No anesthetic agent was used during recording so that synapses would not be depressed. Stimulation. The LDP neurons were identified by antidromic stimulation of the spinal cord at the Ll-L2 level. The stimulating electrode consisted of a pair of flexible wires glued to a plastic strip which was placed beneath the spinal cord in the subdural space. It is called the ventral cord (VC) electrode. Criteria for antidromic responses were those used previously (30). The bipolar electrodes for brain stem stimulation were constructed as reported (25). One was inserted into the left medial longitudinal fasciculus (MLF) and another into the left lateral vestibular nucleus (LVN). The MLF electrode was inserted 5 to 6 mm rostra1 to the obex and 0.5 mm left of midline. Its depth (1.8 to 2.5 mm) was adjusted for lowest threshold (16.2 -+ 5.7 PA) for axonal volleys recorded on the surface of the cervical spinal cord. The LVN electrode was inserted stereotaxically and its position adjusted similarly for lowest threshold (20.2 + 7.1 PA). Another electrode was fixed on the same carrier such that it was in the medial vestibular
464
SKINNER ET AL.
nucleus (MVN) when the LVN electrode was in place. In some experiments an electrode was placed into the left MLF 1 mm caudal to the trochlear nucleus in order to stimulate the tectospinal and interstitiospinal tracts. Its position in the MLF was adjusted so as to produce eye movements at low thresholds (25). All stimuli were constant current pulses of 0.1 ms duration. Testing for supraspinal inputs was with trains of one to three shocks at 333 Hz. Brain stem stimuli for acceptable data did not exceed 100 PA so that activation did not occur beyond 1 mm (1). Natural-type somatosensory stimuli, as used previously (29) consisted of puffs of air for moving hairs, glass probes and calibrated von Frey fibers for touching skin, forceps for pinching skin at pressures from light to noxious, joint movement, and pressure to the deep tissues. Recording. Extracellular spikes were recorded in the cervical enlargement using micropipettes with tips of 4 to 8 pm diameter (1 to 5 MQ) and filled with 4 A4 NaCl or 2 M NaCl saturated with fast green dye. Collision testing between the orthodromic spike evoked by brain stem stimulation and the antidromic spike was always used to show that both spikes originated from the same neuron. A ball-tip wire placed on the dorsal surface of the spinal cord at the recording site (the cord dorsum electrode) was used to record the arrival of the axonal volleys in the descending tracts. Histology. The positions of the stimulating electrodes were ascertained from electrolytic lesions found in frozen sections stained with cresyl violet. The positions of neurons in the spinal cord were reconstructed as reported elsewhere (30). The positions of some neurons were marked with fast green dye iontophoretically ejected from the recording micropipet (33). RESULTS In the 144 long descending propriospinal (LDP) neurons studied in the cervical enlargement, action potentials were orthodromically evoked in 96 (66%) in response to brain stem stimulation. The segmental latency of the response was defined as the shortest time between the peak of the P-wave of the axonal volley near the recording site and the beginning of the action potential. In some trials the stimulus did not elicit an action potential. Latencies ranged from 0.5 to 4.8 ms. A latency of 1.0 ms or less was the criterion for a monosynaptic connection, although some longer latencies also may have been monosynaptic from slower conducting fibers. Using the P-wave of the axonal volley recorded on the surface of the cervical spinal cord as an indication of the arrival of action potentials elicited by brain stem stimulation (3 l), conduction velocities for the faster conducting fibers in the descending tracts were 96 t- 11 m/s (X + SD, N = 25) for MLF fibers and 91 f I 1 m/s for the lateral vestibulospinal tract.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
465
Location of Long Descending Propriospinal Neurons. The left side of segments C5-T 1, ipsilateral to the stimulating electrodes, was searched for LDP neurons, but most were found in C6-C8. Their loci corresponded to those in the ventral horn as reported elsewhere (30). All were in laminae VII and VIII except for one in lateral lamina VI. We concentrated on the ventral group which was in the region of termination of the descending tracts that were stimulated. Medial Longitudinal Fasciculus Stimulation. Stimulation of the MLF elicited spike potentials at monosynaptic latencies (0.5 to 1.0 ms) in 481 144 (33%) of LDP neurons. In 31 of these the spike potential was elicited by one stimulus, 14 required two stimuli, and three required three stimuli. Longer-latency responses were found in an additional 20%, making a total of 53% of LDP neurons which were activated by MLF stimulation. Thresholds for MLF activation ranged from 25 to 100 PA. Figure 1 shows records of a LDP neuron which was monosynaptically activated by single MLF stimuli in A, but was not activated by three stimuli to the LVN in B. A is a composite record of 20 trials in which the neuron responded 12 times at segmental latencies as early as 0.6 ms. In D an antidromic spike from VC stimulation (at dot) is shown, but in C it is missing due to collision with the preceding orthodromic spike elicited by MLF stimulation.
A
MLF
’
MLF+VC
1OOuA
B
3 LVN
190vA
FIG. I. Monosynapticactivation of a LDP neuron by MLF stimulation. A shows 12 responses of a LDP neuron to 20 single stimuli to the MLF at 0.5 Hz. As shown in B it did not respond to a train of three shocks to the LVN at 300 Hz. In D is shown the antidromic spike elicited by a delayed ventral cord (VC) stimulus (at the dot). In C an orthodromic spike evoked by MLF stimulation occurred before the VC stimulus. The antidromic spike from the delayed VC stimulation (at the dot) was blocked by collision with the orthodromic one. The upper traces are microelectrode recordings and the lower traces are cord dorsum recordings. The vertical calibration for all microelectrode recordings is 100 rV/division and the horizontal calibration is I ms/division for A, C, and D and 2 ms/division for B.
466
SKINNER
ET
AL.
This collision test showed that the orthodromic and antidromic spikes originated from the same neuron. Lateral Vestibular Nucleus Stimulation. Monosynaptic responses were elicited in 25 (17%) LDP neurons at segmental latencies of 0.6 to 1.O ms. Twelve of these neurons were activated by one stimulus and the others required two or three stimuli. Longer-latency responses were found in another 33 (23%), making a total of 58 (40%) LDP neurons that were activated by LVN stimulation. Records of the monosynaptic responses of two LDP neurons after LVN stimulation are shown in Fig. 2. The neuron whose responses are depicted in A-C was identified as a LDP cell from its antidromic responses in C to four VC stimuli at 833 Hz. Note the A-B break in the third spike. Figure 2A is a multiple exposure picture of 20 trials of paired stimuli. The first stimulus, which was to the LVN, activated the neuron orthodromically at a monosynaptic latency. The delayed VC stimulus (at dot) elicited an antidromic spike only when the orthodromic one was absent (no collision). Three stimuli to the MLF at 300 Hz in Fig. 2B did not activate this neuron.
A
LVN
lOO,,A+VC
B
3 MLF
lOOup,
C
4 vc
t D
LVN
lOOvA+VC
E;c
-7%
v
F4vch 1
Y -
--
, r”l”ll”
FIG. 2. Brain stem activation of two LDP neurons. The recordings in A-C are from a spontaneously active LDP neuron which responded monosynaptically to LVN stimulation in A. but gave no response to a train of three MLF stimuli at 300 Hz in B. In A there were 20 trials of paired stimuli-first MLF and then VC (at the dot). When there was an orthodromic response to LVN stimulation, the antidromic response elicited by VC stimulation was blocked by collision with the orthodromic spike. When the orthodromic response was absent the antidromic spike was present. Antidromic responses to four VC stimuli at 833 Hz are shown in C. The recordings in D-F are from another LDP neuron which responded at a monosynaptic latency to stimulation of both the LVN (in D) and to the MLF (in E). In D the LVN stimulus was followed by a VC stimulus (at the dot). but the antidromic response was blocked by collision (an unrelated smaller spike is present). Responses to a train of four VC stimuli at 770 Hz are shown in F. Calibration bars apply to all microelectrode traces: 100 @V/division and 1 msjdivision.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
467
Convergence of Medial Longitudinal Fasciculus and Lateral Vestibular Nucleus Inputs. Convergence of input occurred in 39 (27%) LDP neurons. Although convergence was more frequent when latencies were longer than 1 ms, it also occurred at monosynaptic latencies in nine neurons. Figure 2D-F depicts a neuron which was monosynaptically activated by both LVN (D) and MLF (E) stimulation. In D the antidromic spike from the delayed VC stimulus was absent due to collision with the orthodromic spike. Four antidromic responses to VC stimulation at 770 Hz are shown in F. Medial Vestibular Nucleus Stimulation. A stimulating electrode in the medial vestibular nucleus (MVN) ipsilateral to the recording site was used to test for input from the medial vestibulospinal tract in 11 experiments. Only one (l/64) LDP neuron was monosynaptically activated and another five responded at longer latencies. Upper Medial Longitudinal Fasciculus Stimulation. An electrode was placed in the upper-MLF 1 mm caudal to the trochlear nucleus in three experiments in order to stimulate fibers of the tectospinal, interstitiospinal, and other less well known tracts. It was tested at stimulus amplitudes of 300 to 500 PA. None of the 17 LDP neurons tested were monosynaptically activated. The only activation found was at longer latencies (3/ 17). Somatosensory Receptive Fields. The somatosensory receptive fields of 81 LDP neurons were very similar in locus, size, modality, and sign (i.e., excitatory or inhibitory) to those in our previous study (29). We were unable to demonstrate any differences between the receptive fields of monosynaptically activated LDP neurons compared with those activated only at longer latencies or to those which were not activated. Conduction Velocities. The average conduction velocity for all LDP neurons was 67.8 + 10.2 m/s (N = 122). This value is very similar to that found in previous studies (29. 30). DISCUSSION We expected that brain stem stimulation would activate LDP neurons in the C5-Tl segments. In an earlier series of experiments stimulation of the ventral spinal cord at Cl synaptically excited action potentials in 2 l/37 (57%) of LDP neurons (29). Hence, the present series of experiments was conducted with discrete stimulation of brain stem structures in order to extend those earlier results. Also, other investigators have demonstrated monosynaptic activation of propriospinal neurons (of different types) that follows brain stem stimulation (11, 12, 15, 35). Brain stem stimulation sites were chosen in order to activate several well known descending pathways terminating in the ventral horn. The lateral vestibulospinal tract was stimulated at its origin in the LVN. The MLF was
468
SKINNER
ET AL.
stimulated 5 to 6 mm rostral to the obex (8, 34). At that site the MLF contains reticulospinal, interstitiospinal, medial vestibulospinal, and tectospinal fibers (4) and possibly other fibers originating in the upper midbrain and hypothalamus. The tectospinal tract terminates mainly in the upper cervical spinal cord, rostra1 to our recording region. The medial vestibulospinal tract descends bilaterally as far as the upper thoracic region and the interstitiospinal tract descends mainly ipsilaterally to all regions of the spinal cord. Effects of these latter three tracts and other fibers with a more rostra1 origin were tested with electrodes in the medial vestibular nucleus and in the upper part of the MLF about 1 mm caudal to the trochlear nucleus. Input from MLF fibers to LDP neurons was observed more frequently than input from the LVN. Of the 64 LDP neurons that were activated monosynaptically by MLF or LVN stimulation, about two-thirds of these were activated by MLF stimulation and one-third by LVN. and some by both of these. A similar result with regard to the relative frequency of input for MLF versus LVN stimulation was found in limb motor neurons (8, 34). Another 22% of LDP neurons responded only at longer latencies, with MLF input again being more effective. Some of those responding at latencies longer than 1 ms also could have had monosynaptic connections. Other LDP neurons could have been excited subliminally, or inhibited, without being detected by extracellular recording. Preliminary unpublished evidence from intracellular recording in LDP neurons shows that there are monosynaptic excitatory and longer-latency inhibitory postsynaptic potentials after MLF or LVN stimulation. Stimulation of the MVN and upper MLF was less effective. About half of the LDP neurons which were activated at monosynaptic latencies received convergent input from both MLF and LVN-mostly at longer latencies. Convergence was less frequent in LDP neurons activated only at longer latencies. It appears that convergence of descending input is common in propriospinal (PS) neurons. Vasilenko and Kostyukov (35) found short descending PS neurons in the ventromedial region of the L4L5 segments which received substantial convergence from reticula- and vestibulospinal fibers. They observed monosynaptic excitatory postsynaptic potentials from reticulospinal fibers in 15/28 cells, but only polysynaptic effects after LVN stimulation. Also, many PS neurons in C3-C4 were monosynaptically excited by cortico-, rubro-, and reticulospinal fibers (11, 12). We reported elsewhere that 16% of the LDP neurons have an ascending axonal branch which was stimulated at Cl (29). This type of neuron, therefore, is both an ascending tract cell and an LDP cell. Mannen (18) showed anatomically that spinal neurons in the lumbosacral region can possess ascending and descending axons of comparable size. The Goteborg
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
469
group discovered similar propriospinal neurons in C3-C4 which have an ascending branch to the lateral reticular nucleus and a descending branch mainly to forelimb motor neurons. However, an ascending axonal branch was found in only 11% of those projecting more caudally than T9 (2) and the ascending axon generally had a slower conduction velocity. Hirai et al. (9) reported that 43% of the ventral group of spinocerebellar neurons originating in C7-Tl possessed a collateral axon which descended as far as T13. It appears then, that different populations of PS neurons vary as to the number which project to the brain stem. The LDP group as a whole, however, probably has a small percentage (about 20%) of neurons with ascending axons. It is possible that activation of some LDP neurons occurred via antidromic activation of fibers ascending to the brain stem which have collateral axons projecting to LDP cells. We believe that such a pathway is unlikely to have produced the observed effects for two reasons. First, the positions of the stimulating electrodes in the MLF do not correspond well to the loci of spinoreticular afferent fibers described by Rossi and Brodal (26). They reported that the densest terminations of such afferent fibers were found bilaterally in the nucleus reticularis (Nr) pontis caudalis and ipsilaterally in the ventral parts of the Nr gigantocellularis and Nr ventralis. Thus, by means of careful placement of the MLF electrodes and the restriction of stimulating currents to 100 PA, few afferent fibers should have activated by these electrodes. Second, the conduction velocities of ascending fibers are on the average slower than those of descending fibers (6, 20). Therefore, the observed responses were most likely due to the faster conducting fibers in descending tracts (1, 24), although some responses could have occurred via collaterals of ascending fibers. The possibility also exists that some of the observed effects were due to stimulation of collaterals of descending tract fibers projecting near the stimulating electrodes. Because the projection of the LVN to the medullary reticular formation is sparse (14) it is unlikely that the MLF electrode stimulated a significant number of collaterals of lateral vestibulospinal tract fibers. However, the Nr gigantocellularis and the Nr pontis caudalis projections to the LVN must be considered (10). If some of those projections are collaterals of reticulospinal neurons, then some effects elicited by LVN stimulation could actually be due to reticulospinal neurons. This is particularly true for those LDP neurons activated monosynaptically by both LVN and MLF stimulation. However, there were a significant number of instances (16/25) in which monosynaptic activation from LVN stimulation occurred when there was no monosynaptic activation by MLF stimulation, even when tested to 160 PA. Therefore, for the above reasons and the fact that the upper-MLF and MVN electrodes were not very effective, it is most
470
SKINNER
ET AL.
likely that the responses evoked by MLF stimulation were from reticulospinal fibers originating in the medullary and/or pontine reticular formation and that most responses evoked by LVN stimulation were from lateral vestibulospinal tract fibers. Accumulated evidence indicates that LDP neurons can modulate the motor behavior of the hind limbs and that appropriate anatomic and physiologic substrates for this function exist. Shenington and Laslett (28) demonstrated several complex reflexes involving the hind limbs of dogs (spinalized at C5-6) which were initiated by afferent fibers in the forelimbs. Subsequent electrophysiological studies indicated that LDP neurons are involved in this reflex pathway (16, 27, 29). Anatomic studies revealed that these LDP projections terminate in the lumbosacral enlargement bilaterally (3, 19, 30). Intracellular recording showed that there are monosynaptic excitatory connections from LDP neurons with many lumbosacral motor neurons and interneurons (13). The complex nature of the forelimb-hind limb reflexes presumably is mediated by subpopulations of LDP neurons. Our finding that LDP neurons can be modulated by monosynaptic supraspinal input suggests a role for LDP neurons in transmitting commands from the brain stem, as well as information from forelimb afferent fibers, to the hind limb. REFERENCES C., M. MAEDA, B. W. PETERSON, AND V. J. WILSON. 1974. Cervical branching of lumbar vestibulospinal axons. J. Physiol. (London) 243: 499-522. ALSTERMARK, B., S. LINDSTROM, A. LUNDBERG, AND E. SYBIRSKA. 1981. Integration in descending motor pathways controlling the forelimb in the cat. 8. Ascending projection to the lateral reticular nucleus from C3-C4 propriospinal neurones also projecting to forelimb motoneurones. Exp. Brain Rex 42: 282-298. BARILARI, M. G., AND H. G. J. M. KUYPERS. 1969. Propriospinal fibers interconnecting the spinal enlargements in the cat. Bruin Rex 14: 321-330. BRODAL, A. 198 I. Neurological Anatomy, 3rd ed. Oxford Univ. Press, New York/Oxford. DELWAIDE, P. J.. C. FIGIEL, AND C. RICHELLE. 1977. Effects of postural changes of the upper limb on reflex transmission in the lower limb. Cervicolumbar reflex interactions in man. J. Neural. Neurosurg. Psychiatry 40: 6 16-62 1. FIELDS, H. L., C. H. CLANTON, AND S. D. ANDERSON. 1977. Somatosensory properties of spinoreticular neurons in the cat. Brain Rex 120: 49-66. GRILLNER, S. 1975. Locomotion in vertebrates: central mechanisms and reflex interaction.
1. ABZUG, 2.
3, 4. 5. 6. 7.
Physiol.
Rev. 55: 247-304.
8. GRILLNER, S.. T. HONGO, AND S. LUND. 1971. Convergent effects on alpha motoneurones from the vestibulospinal tract and a pathway descending in the medial longitudinal fasciculus. Exp. Brain. Rex 12: 457-479. 9. HIRAI, N., T. HONGO, AND Y. YAMAGUCHI. 1978. Spinocerebellar tract neurones with long descending axon collaterals. Brain Res. 142: l47- I5 1. 10. HODDEVIK, G. H., A. BRODAL, AND F. WALBERG. 1975. The reticulovestibular projection in the cat. An experimental study with silver impregnation methods. Bruin Rex 94: 383-399.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
471
I I. ILLERT, M., E. JANKOWSKA, A. LUNDBERG, AND A. ODUTOLA. 1981. Integration in descending motor pathways controlling the forelimb in the cat. 7. Effects from the reticular formation on C3-C4 propriospinal neurones. E,rp. Bruin Res. 42: 269-28 I. 12. ILLERT, M., A. LUNDBERG, Y. PADEL, AND R. TANAKA. 1978. Integration in descending motor pathways controlling the forelimb in the cat. 5. Properties of and monosynaptic excitatory convergence on C3-C4 propriospinal neurones. &p. Bruin Rcs. 33: IOl130. 13. JANKOWSKA,E., A. LUNDBERG, W. J. ROBERTS, AND D. STUART. 1974. A long propriospinal system with direct effect on motoneurones and on intemeurones in the cat lumbosacral cord. Exp. Brain Res. 21: 169-194. 14. LADPLI, R., AND A. BRODAL. 1968. Experimental studies of commissural and reticular formation projections from the vestibular nuclei in the cat. Bruin Rex 8: 65-96. 15. LLOYD, D. P. C. 1941. Activity in neurons of the bulbospinal correlating system. J. Neurophysiol. 4: 11S- 134. 16. LLOYD, D. P. C. 1942. Mediation of descending long spinal reflex activity. J. Neurophysiol. 5: 435-458.
17. LLOYD, D. P. C., AND A. K. MCINTYRE. 1948. Analysis of forelimb-hindlimb reflex activity in acutely decapitate cats. J. Neurophysiol. 11: 455-470. 18. MANNEN, H. 1975. Reconstruction of axonal trajectory of individual neurons in the spinal cord using Golgi-stained serial sections. J. Comp. Neurol. 159: 357-374. 19. MATSUSHITA, M., M. IKEDA, AND Y. HOSOYA. 1979. The location of spinal neurons with long descending axons (long descending propriospinal tract neurons) in the cat: a study with the horseradish peroxidase technique. J. Camp. Neurol. 184: 63-80. 20. MAUNZ, R. A., N. G. PITTS, AND B. W. PETERSON. 1978. Cat spinoreticular neurons: locations, responses and changes in responses during repetitive stimulation. Bruin Res. 148: 365-379.
21. MEINCK, H.-M., AND B. PIESIUR-STREHLOW. 1981. Reflexes evoked in leg muscles from arm afferents: a propriospinal pathway in man? E.rp. Brain Rex 43: 78-86. 22. MILLER, S., AND J. VAN DER BERG. 1973. The function of long propriospinal pathways in the co-ordination of quadrupedal stepping in the cat. Pages 561-577 in R. B. STEIN, K. G. PEARSON, R. S. SMITH, AND J. B. REDFORD, Eds., Control aJ‘ Posture and Locomotion. (Advances in Behavioral Biology. Vol. 7). Plenum. New York/London. 23. MOLENAAR, I., AND H. G. J. M. KUYPERS. 1978. Cells of origin of propriospinal fibers and of fibers ascending to supraspinal levels. A HRP study in cat and rhesus monkey. Brain
Rex
152: 429-450.
24. PETERSON, B. W., R. A. MAUNZ, N. G. PITTS, AND R. G. MACKEL. 1975. Patterns of projection and branching of reticulospinal neurons. Exp. Brain Res. 23: 333-35 I. 25. REMMEL, R. S., R. D. SKINNER, AND L. B. MINOR. 1980. Eighth nerve activation of cat pontine reticular neurons which project in or near the ascending medial longitudinal fasciculus. Exp. Neurol. 70: 706-7 1I. 26. ROSSI,G. F., AND A. BRODAL. 1957. Terminal distribution of spinoreticular fibers in the cat. Arch. Neurol. Psychiatry (Chicago) 78: 439-453. 27. SCHOMBURG, E. D.. H.-M. MEINCK, J. HAUSTEIN, AND J. ROESLER. 1978. Functional organization of the spinal reflex pathways from forelimb afferents to hindlimb motorneurones in the cat. Brain Res. 139: 2 l-33. 28. SHERRINGTON,C. S.. AND E. E. LASLETT. 1903. Observations on some spinal reflexes and the interconnection of spinal segments. J. Physiol. (London) 29: 58-96. 29. SKINNER, R. D., R. J. ADAMS, AND R. S. REMMEL. 1980. Responses of long descending propriospinal neurons to natural and electrical types of stimuli in cat. Brain Res. 196: 387-403.
472
SKINNER
ET AL.
30. SKINNER, R. D., J. D. COULTER, R. J. ADAMS, AND R. S. REMMEL. 1979. Cells of origin of long descending propriospinal fibers connecting the spinal enlargements in cat and monkey determined by horseradish peroxidase and electrophysiological techniques. J. Comp.
Neural
188: 443-454.
31. SKINNER, R. D., AND R. S. REMMEL. 1978. Monosynaptic inputs to lumbar intemeurons from the lateral vestibulospinal tract and the medial longitudinal fasciculus. Neurosci. Lett. 10: 259-264.
32. SKINNER, R. D., AND R. S. REMMEL. 1980. Monosynaptic supraspinal inputs to long descending propriospinal neurons in cats. Sot. Neurosci. Abstr. 6: 24. 33. THOMAS, R. C., AND V. J. WILSON, 1965. Precise localization of Renshaw cells with a new marking technique. Nature (London) 206: 2 1 l-2 13. 34. WILSON, V. J., AND M. YOSHIDA. 1969. Comparison of effects of stimulation of Deiters’ nucleus and medial longitudinal fasciculus on neck, forelimb, and hindlimb motoneurons. J. Neurophysiol.
32: 743-758.
35. VASILENKO, D. A., AND A. I. KOS~YUKOV. 1976. Brain stem and primary afferent projections to the ventromedial group of propriospinal neurones in the cat. Brain Res. 117: 141-146. 36. YEZIERSKI, R. P., J. L. CULBERSON, AND P. B. BROWN. 1980. Cells of origin of propriospinal connections to cat lumbosacral gray as determined with horseradish peroxidase. Exp. Neural. 69: 493-5 12.
NEUROLOGY
86.462-472
(1984)
Monosynaptic Activation of Long Descending Propriospinal Neurons from the Lateral Vestibular Nucleus and the Medial Longitudinal Fasciculus R. D. SKINNER, Department
R. S. REMMEL,
AND L. B. MINOR’
of Anatomy and Department of Physiology and Biophysics Arkansas for Medical Sciences. Little Rock, Arkansas Received
March
27, 1984: revision
received
(R.S.R.), 72205
University
of
July 27, I984
In decerebrate cats long descending propriospinal (LDP) neurons were recorded extracellularly in the cervical enlargement. They were identified antidromically by spinal cord stimulation at the Ll -L2 level. Inputs to these cells were tested by stimulating the medial longitudinal fasciculus (MLF) 5 to 6 mm rostra1 to the obex. the lateral vestibular nucleus (LVN). the upper MLF I mm caudal to the trochlear nucleus, and the medial vestibular nucleus (MVN), all on the ipsilateral side. Action potentials were elicited in 44% (64/144) of LDP neurons in the ventral horn (laminae VII, VIII) at a segmental latency of 1 ms or less following brain stem stimulation. This was considered to be a monosynaptic latency. The most effective stimulation sites were the MLF and the LVN. MLF stimulation accounted for about two-thirds of the monosynaptically elicited action potentials and LVN for about one-third. Another 22% of LDP neurons responded at longer latencies, but some of those responses may also have been monosynaptic. Stimulation of the upper MLF and the MVN were much less effective,indicating that the MLF input was predominantly from fibers originating in the medullary and/or pontine reticular formation. Q 1984 Academic
Press. Inc.
INTRODUCTION Propriospinal neurons connecting different segments of the spinal cord constitute a heterogeneous population of neurons having different lengths Abbreviations: LDP-long descending propriospinal: MLF-medial longitudinal fasciculus: LVN, MVN-lateral, medial vestibular nucleus; Nr-nucleus reticularis; VC-ventral cord; PS-propriospinal. ’ We thank Nancy Stone for typing and Gael Sammartino for technical assistance. David Borne participated in some of the experiments. Lloyd B. Minor was an extem from Brown University, Division of Biology and Medicine. His present address is Department of Surgery, Duke University School of Medicine. Durham, NC 27705. This work was sponsored by U.S. Public Health Service grant NS 10304 and by National Science Foundation grant ISP 801 1447. 462 OOl4-4886184 $3.00 Copyright 0 1984 by Academic Press. Inc. All n&s of rcproductron I” any form rescrred.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
463
of axons and, most likely, a variety of functions. The lengths of their axons vary from one segment to possibly the length of the spinal cord (19, 23, 30, 36). They function in spinal reflexes in cats (15-l 7, 22, 27), in dogs (28), and probably in man (5, 21). An intriguing, but untested, hypothesis of the function of long propriospinal neurons is that they are part of the spinal locomotion generators and help to coordinate the rostra1 and caudal pairs of limbs during locomotion in tetrapods (7, 22). The purpose of this study was to extend previous observations regarding descending input to long descending propriospinal (LDP) neurons. The LDP neurons studied had their cell bodies in the cervical enlargement and projected to the lumbosacral region (30). Activation at short latencies, particularly those short enough for a monosynaptic connection, would indicate that these cells mediate directly the effects of some descending pathways. A brief description of this work has been reported (32). METHODS In 25 cats under ether anesthesia, the carotid arteries were ligated and a precollicular decerebration was done. Blood pressure was maintained above 90 mm Hg by means of an infusion of norepinephrine (80 pg/ml). Rectal temperature and end-tidal Pcoz were monitored and maintained near normal values. The occipital bone and midline cerebellum were removed to expose the fourth ventricle. Laminectomies were made at C5-Tl and Ll -L2. The cat’s head was held in a stereotaxic frame and the T2 spinous process was rigidly clamped. The cat was paralyzed with gallamine triethiodide and artificially ventilated. A bilateral pneumothorax reduced movements. No anesthetic agent was used during recording so that synapses would not be depressed. Stimulation. The LDP neurons were identified by antidromic stimulation of the spinal cord at the Ll-L2 level. The stimulating electrode consisted of a pair of flexible wires glued to a plastic strip which was placed beneath the spinal cord in the subdural space. It is called the ventral cord (VC) electrode. Criteria for antidromic responses were those used previously (30). The bipolar electrodes for brain stem stimulation were constructed as reported (25). One was inserted into the left medial longitudinal fasciculus (MLF) and another into the left lateral vestibular nucleus (LVN). The MLF electrode was inserted 5 to 6 mm rostra1 to the obex and 0.5 mm left of midline. Its depth (1.8 to 2.5 mm) was adjusted for lowest threshold (16.2 -+ 5.7 PA) for axonal volleys recorded on the surface of the cervical spinal cord. The LVN electrode was inserted stereotaxically and its position adjusted similarly for lowest threshold (20.2 + 7.1 PA). Another electrode was fixed on the same carrier such that it was in the medial vestibular
464
SKINNER ET AL.
nucleus (MVN) when the LVN electrode was in place. In some experiments an electrode was placed into the left MLF 1 mm caudal to the trochlear nucleus in order to stimulate the tectospinal and interstitiospinal tracts. Its position in the MLF was adjusted so as to produce eye movements at low thresholds (25). All stimuli were constant current pulses of 0.1 ms duration. Testing for supraspinal inputs was with trains of one to three shocks at 333 Hz. Brain stem stimuli for acceptable data did not exceed 100 PA so that activation did not occur beyond 1 mm (1). Natural-type somatosensory stimuli, as used previously (29) consisted of puffs of air for moving hairs, glass probes and calibrated von Frey fibers for touching skin, forceps for pinching skin at pressures from light to noxious, joint movement, and pressure to the deep tissues. Recording. Extracellular spikes were recorded in the cervical enlargement using micropipettes with tips of 4 to 8 pm diameter (1 to 5 MQ) and filled with 4 A4 NaCl or 2 M NaCl saturated with fast green dye. Collision testing between the orthodromic spike evoked by brain stem stimulation and the antidromic spike was always used to show that both spikes originated from the same neuron. A ball-tip wire placed on the dorsal surface of the spinal cord at the recording site (the cord dorsum electrode) was used to record the arrival of the axonal volleys in the descending tracts. Histology. The positions of the stimulating electrodes were ascertained from electrolytic lesions found in frozen sections stained with cresyl violet. The positions of neurons in the spinal cord were reconstructed as reported elsewhere (30). The positions of some neurons were marked with fast green dye iontophoretically ejected from the recording micropipet (33). RESULTS In the 144 long descending propriospinal (LDP) neurons studied in the cervical enlargement, action potentials were orthodromically evoked in 96 (66%) in response to brain stem stimulation. The segmental latency of the response was defined as the shortest time between the peak of the P-wave of the axonal volley near the recording site and the beginning of the action potential. In some trials the stimulus did not elicit an action potential. Latencies ranged from 0.5 to 4.8 ms. A latency of 1.0 ms or less was the criterion for a monosynaptic connection, although some longer latencies also may have been monosynaptic from slower conducting fibers. Using the P-wave of the axonal volley recorded on the surface of the cervical spinal cord as an indication of the arrival of action potentials elicited by brain stem stimulation (3 l), conduction velocities for the faster conducting fibers in the descending tracts were 96 t- 11 m/s (X + SD, N = 25) for MLF fibers and 91 f I 1 m/s for the lateral vestibulospinal tract.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
465
Location of Long Descending Propriospinal Neurons. The left side of segments C5-T 1, ipsilateral to the stimulating electrodes, was searched for LDP neurons, but most were found in C6-C8. Their loci corresponded to those in the ventral horn as reported elsewhere (30). All were in laminae VII and VIII except for one in lateral lamina VI. We concentrated on the ventral group which was in the region of termination of the descending tracts that were stimulated. Medial Longitudinal Fasciculus Stimulation. Stimulation of the MLF elicited spike potentials at monosynaptic latencies (0.5 to 1.0 ms) in 481 144 (33%) of LDP neurons. In 31 of these the spike potential was elicited by one stimulus, 14 required two stimuli, and three required three stimuli. Longer-latency responses were found in an additional 20%, making a total of 53% of LDP neurons which were activated by MLF stimulation. Thresholds for MLF activation ranged from 25 to 100 PA. Figure 1 shows records of a LDP neuron which was monosynaptically activated by single MLF stimuli in A, but was not activated by three stimuli to the LVN in B. A is a composite record of 20 trials in which the neuron responded 12 times at segmental latencies as early as 0.6 ms. In D an antidromic spike from VC stimulation (at dot) is shown, but in C it is missing due to collision with the preceding orthodromic spike elicited by MLF stimulation.
A
MLF
’
MLF+VC
1OOuA
B
3 LVN
190vA
FIG. I. Monosynapticactivation of a LDP neuron by MLF stimulation. A shows 12 responses of a LDP neuron to 20 single stimuli to the MLF at 0.5 Hz. As shown in B it did not respond to a train of three shocks to the LVN at 300 Hz. In D is shown the antidromic spike elicited by a delayed ventral cord (VC) stimulus (at the dot). In C an orthodromic spike evoked by MLF stimulation occurred before the VC stimulus. The antidromic spike from the delayed VC stimulation (at the dot) was blocked by collision with the orthodromic one. The upper traces are microelectrode recordings and the lower traces are cord dorsum recordings. The vertical calibration for all microelectrode recordings is 100 rV/division and the horizontal calibration is I ms/division for A, C, and D and 2 ms/division for B.
466
SKINNER
ET
AL.
This collision test showed that the orthodromic and antidromic spikes originated from the same neuron. Lateral Vestibular Nucleus Stimulation. Monosynaptic responses were elicited in 25 (17%) LDP neurons at segmental latencies of 0.6 to 1.O ms. Twelve of these neurons were activated by one stimulus and the others required two or three stimuli. Longer-latency responses were found in another 33 (23%), making a total of 58 (40%) LDP neurons that were activated by LVN stimulation. Records of the monosynaptic responses of two LDP neurons after LVN stimulation are shown in Fig. 2. The neuron whose responses are depicted in A-C was identified as a LDP cell from its antidromic responses in C to four VC stimuli at 833 Hz. Note the A-B break in the third spike. Figure 2A is a multiple exposure picture of 20 trials of paired stimuli. The first stimulus, which was to the LVN, activated the neuron orthodromically at a monosynaptic latency. The delayed VC stimulus (at dot) elicited an antidromic spike only when the orthodromic one was absent (no collision). Three stimuli to the MLF at 300 Hz in Fig. 2B did not activate this neuron.
A
LVN
lOO,,A+VC
B
3 MLF
lOOup,
C
4 vc
t D
LVN
lOOvA+VC
E;c
-7%
v
F4vch 1
Y -
--
, r”l”ll”
FIG. 2. Brain stem activation of two LDP neurons. The recordings in A-C are from a spontaneously active LDP neuron which responded monosynaptically to LVN stimulation in A. but gave no response to a train of three MLF stimuli at 300 Hz in B. In A there were 20 trials of paired stimuli-first MLF and then VC (at the dot). When there was an orthodromic response to LVN stimulation, the antidromic response elicited by VC stimulation was blocked by collision with the orthodromic spike. When the orthodromic response was absent the antidromic spike was present. Antidromic responses to four VC stimuli at 833 Hz are shown in C. The recordings in D-F are from another LDP neuron which responded at a monosynaptic latency to stimulation of both the LVN (in D) and to the MLF (in E). In D the LVN stimulus was followed by a VC stimulus (at the dot). but the antidromic response was blocked by collision (an unrelated smaller spike is present). Responses to a train of four VC stimuli at 770 Hz are shown in F. Calibration bars apply to all microelectrode traces: 100 @V/division and 1 msjdivision.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
467
Convergence of Medial Longitudinal Fasciculus and Lateral Vestibular Nucleus Inputs. Convergence of input occurred in 39 (27%) LDP neurons. Although convergence was more frequent when latencies were longer than 1 ms, it also occurred at monosynaptic latencies in nine neurons. Figure 2D-F depicts a neuron which was monosynaptically activated by both LVN (D) and MLF (E) stimulation. In D the antidromic spike from the delayed VC stimulus was absent due to collision with the orthodromic spike. Four antidromic responses to VC stimulation at 770 Hz are shown in F. Medial Vestibular Nucleus Stimulation. A stimulating electrode in the medial vestibular nucleus (MVN) ipsilateral to the recording site was used to test for input from the medial vestibulospinal tract in 11 experiments. Only one (l/64) LDP neuron was monosynaptically activated and another five responded at longer latencies. Upper Medial Longitudinal Fasciculus Stimulation. An electrode was placed in the upper-MLF 1 mm caudal to the trochlear nucleus in three experiments in order to stimulate fibers of the tectospinal, interstitiospinal, and other less well known tracts. It was tested at stimulus amplitudes of 300 to 500 PA. None of the 17 LDP neurons tested were monosynaptically activated. The only activation found was at longer latencies (3/ 17). Somatosensory Receptive Fields. The somatosensory receptive fields of 81 LDP neurons were very similar in locus, size, modality, and sign (i.e., excitatory or inhibitory) to those in our previous study (29). We were unable to demonstrate any differences between the receptive fields of monosynaptically activated LDP neurons compared with those activated only at longer latencies or to those which were not activated. Conduction Velocities. The average conduction velocity for all LDP neurons was 67.8 + 10.2 m/s (N = 122). This value is very similar to that found in previous studies (29. 30). DISCUSSION We expected that brain stem stimulation would activate LDP neurons in the C5-Tl segments. In an earlier series of experiments stimulation of the ventral spinal cord at Cl synaptically excited action potentials in 2 l/37 (57%) of LDP neurons (29). Hence, the present series of experiments was conducted with discrete stimulation of brain stem structures in order to extend those earlier results. Also, other investigators have demonstrated monosynaptic activation of propriospinal neurons (of different types) that follows brain stem stimulation (11, 12, 15, 35). Brain stem stimulation sites were chosen in order to activate several well known descending pathways terminating in the ventral horn. The lateral vestibulospinal tract was stimulated at its origin in the LVN. The MLF was
468
SKINNER
ET AL.
stimulated 5 to 6 mm rostral to the obex (8, 34). At that site the MLF contains reticulospinal, interstitiospinal, medial vestibulospinal, and tectospinal fibers (4) and possibly other fibers originating in the upper midbrain and hypothalamus. The tectospinal tract terminates mainly in the upper cervical spinal cord, rostra1 to our recording region. The medial vestibulospinal tract descends bilaterally as far as the upper thoracic region and the interstitiospinal tract descends mainly ipsilaterally to all regions of the spinal cord. Effects of these latter three tracts and other fibers with a more rostra1 origin were tested with electrodes in the medial vestibular nucleus and in the upper part of the MLF about 1 mm caudal to the trochlear nucleus. Input from MLF fibers to LDP neurons was observed more frequently than input from the LVN. Of the 64 LDP neurons that were activated monosynaptically by MLF or LVN stimulation, about two-thirds of these were activated by MLF stimulation and one-third by LVN. and some by both of these. A similar result with regard to the relative frequency of input for MLF versus LVN stimulation was found in limb motor neurons (8, 34). Another 22% of LDP neurons responded only at longer latencies, with MLF input again being more effective. Some of those responding at latencies longer than 1 ms also could have had monosynaptic connections. Other LDP neurons could have been excited subliminally, or inhibited, without being detected by extracellular recording. Preliminary unpublished evidence from intracellular recording in LDP neurons shows that there are monosynaptic excitatory and longer-latency inhibitory postsynaptic potentials after MLF or LVN stimulation. Stimulation of the MVN and upper MLF was less effective. About half of the LDP neurons which were activated at monosynaptic latencies received convergent input from both MLF and LVN-mostly at longer latencies. Convergence was less frequent in LDP neurons activated only at longer latencies. It appears that convergence of descending input is common in propriospinal (PS) neurons. Vasilenko and Kostyukov (35) found short descending PS neurons in the ventromedial region of the L4L5 segments which received substantial convergence from reticula- and vestibulospinal fibers. They observed monosynaptic excitatory postsynaptic potentials from reticulospinal fibers in 15/28 cells, but only polysynaptic effects after LVN stimulation. Also, many PS neurons in C3-C4 were monosynaptically excited by cortico-, rubro-, and reticulospinal fibers (11, 12). We reported elsewhere that 16% of the LDP neurons have an ascending axonal branch which was stimulated at Cl (29). This type of neuron, therefore, is both an ascending tract cell and an LDP cell. Mannen (18) showed anatomically that spinal neurons in the lumbosacral region can possess ascending and descending axons of comparable size. The Goteborg
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
469
group discovered similar propriospinal neurons in C3-C4 which have an ascending branch to the lateral reticular nucleus and a descending branch mainly to forelimb motor neurons. However, an ascending axonal branch was found in only 11% of those projecting more caudally than T9 (2) and the ascending axon generally had a slower conduction velocity. Hirai et al. (9) reported that 43% of the ventral group of spinocerebellar neurons originating in C7-Tl possessed a collateral axon which descended as far as T13. It appears then, that different populations of PS neurons vary as to the number which project to the brain stem. The LDP group as a whole, however, probably has a small percentage (about 20%) of neurons with ascending axons. It is possible that activation of some LDP neurons occurred via antidromic activation of fibers ascending to the brain stem which have collateral axons projecting to LDP cells. We believe that such a pathway is unlikely to have produced the observed effects for two reasons. First, the positions of the stimulating electrodes in the MLF do not correspond well to the loci of spinoreticular afferent fibers described by Rossi and Brodal (26). They reported that the densest terminations of such afferent fibers were found bilaterally in the nucleus reticularis (Nr) pontis caudalis and ipsilaterally in the ventral parts of the Nr gigantocellularis and Nr ventralis. Thus, by means of careful placement of the MLF electrodes and the restriction of stimulating currents to 100 PA, few afferent fibers should have activated by these electrodes. Second, the conduction velocities of ascending fibers are on the average slower than those of descending fibers (6, 20). Therefore, the observed responses were most likely due to the faster conducting fibers in descending tracts (1, 24), although some responses could have occurred via collaterals of ascending fibers. The possibility also exists that some of the observed effects were due to stimulation of collaterals of descending tract fibers projecting near the stimulating electrodes. Because the projection of the LVN to the medullary reticular formation is sparse (14) it is unlikely that the MLF electrode stimulated a significant number of collaterals of lateral vestibulospinal tract fibers. However, the Nr gigantocellularis and the Nr pontis caudalis projections to the LVN must be considered (10). If some of those projections are collaterals of reticulospinal neurons, then some effects elicited by LVN stimulation could actually be due to reticulospinal neurons. This is particularly true for those LDP neurons activated monosynaptically by both LVN and MLF stimulation. However, there were a significant number of instances (16/25) in which monosynaptic activation from LVN stimulation occurred when there was no monosynaptic activation by MLF stimulation, even when tested to 160 PA. Therefore, for the above reasons and the fact that the upper-MLF and MVN electrodes were not very effective, it is most
470
SKINNER
ET AL.
likely that the responses evoked by MLF stimulation were from reticulospinal fibers originating in the medullary and/or pontine reticular formation and that most responses evoked by LVN stimulation were from lateral vestibulospinal tract fibers. Accumulated evidence indicates that LDP neurons can modulate the motor behavior of the hind limbs and that appropriate anatomic and physiologic substrates for this function exist. Shenington and Laslett (28) demonstrated several complex reflexes involving the hind limbs of dogs (spinalized at C5-6) which were initiated by afferent fibers in the forelimbs. Subsequent electrophysiological studies indicated that LDP neurons are involved in this reflex pathway (16, 27, 29). Anatomic studies revealed that these LDP projections terminate in the lumbosacral enlargement bilaterally (3, 19, 30). Intracellular recording showed that there are monosynaptic excitatory connections from LDP neurons with many lumbosacral motor neurons and interneurons (13). The complex nature of the forelimb-hind limb reflexes presumably is mediated by subpopulations of LDP neurons. Our finding that LDP neurons can be modulated by monosynaptic supraspinal input suggests a role for LDP neurons in transmitting commands from the brain stem, as well as information from forelimb afferent fibers, to the hind limb. REFERENCES C., M. MAEDA, B. W. PETERSON, AND V. J. WILSON. 1974. Cervical branching of lumbar vestibulospinal axons. J. Physiol. (London) 243: 499-522. ALSTERMARK, B., S. LINDSTROM, A. LUNDBERG, AND E. SYBIRSKA. 1981. Integration in descending motor pathways controlling the forelimb in the cat. 8. Ascending projection to the lateral reticular nucleus from C3-C4 propriospinal neurones also projecting to forelimb motoneurones. Exp. Brain Rex 42: 282-298. BARILARI, M. G., AND H. G. J. M. KUYPERS. 1969. Propriospinal fibers interconnecting the spinal enlargements in the cat. Bruin Rex 14: 321-330. BRODAL, A. 198 I. Neurological Anatomy, 3rd ed. Oxford Univ. Press, New York/Oxford. DELWAIDE, P. J.. C. FIGIEL, AND C. RICHELLE. 1977. Effects of postural changes of the upper limb on reflex transmission in the lower limb. Cervicolumbar reflex interactions in man. J. Neural. Neurosurg. Psychiatry 40: 6 16-62 1. FIELDS, H. L., C. H. CLANTON, AND S. D. ANDERSON. 1977. Somatosensory properties of spinoreticular neurons in the cat. Brain Rex 120: 49-66. GRILLNER, S. 1975. Locomotion in vertebrates: central mechanisms and reflex interaction.
1. ABZUG, 2.
3, 4. 5. 6. 7.
Physiol.
Rev. 55: 247-304.
8. GRILLNER, S.. T. HONGO, AND S. LUND. 1971. Convergent effects on alpha motoneurones from the vestibulospinal tract and a pathway descending in the medial longitudinal fasciculus. Exp. Brain. Rex 12: 457-479. 9. HIRAI, N., T. HONGO, AND Y. YAMAGUCHI. 1978. Spinocerebellar tract neurones with long descending axon collaterals. Brain Res. 142: l47- I5 1. 10. HODDEVIK, G. H., A. BRODAL, AND F. WALBERG. 1975. The reticulovestibular projection in the cat. An experimental study with silver impregnation methods. Bruin Rex 94: 383-399.
MONOSYNAPTIC
ACTIVATION
OF
LDP
NEURONS
471
I I. ILLERT, M., E. JANKOWSKA, A. LUNDBERG, AND A. ODUTOLA. 1981. Integration in descending motor pathways controlling the forelimb in the cat. 7. Effects from the reticular formation on C3-C4 propriospinal neurones. E,rp. Bruin Res. 42: 269-28 I. 12. ILLERT, M., A. LUNDBERG, Y. PADEL, AND R. TANAKA. 1978. Integration in descending motor pathways controlling the forelimb in the cat. 5. Properties of and monosynaptic excitatory convergence on C3-C4 propriospinal neurones. &p. Bruin Rcs. 33: IOl130. 13. JANKOWSKA,E., A. LUNDBERG, W. J. ROBERTS, AND D. STUART. 1974. A long propriospinal system with direct effect on motoneurones and on intemeurones in the cat lumbosacral cord. Exp. Brain Res. 21: 169-194. 14. LADPLI, R., AND A. BRODAL. 1968. Experimental studies of commissural and reticular formation projections from the vestibular nuclei in the cat. Bruin Rex 8: 65-96. 15. LLOYD, D. P. C. 1941. Activity in neurons of the bulbospinal correlating system. J. Neurophysiol. 4: 11S- 134. 16. LLOYD, D. P. C. 1942. Mediation of descending long spinal reflex activity. J. Neurophysiol. 5: 435-458.
17. LLOYD, D. P. C., AND A. K. MCINTYRE. 1948. Analysis of forelimb-hindlimb reflex activity in acutely decapitate cats. J. Neurophysiol. 11: 455-470. 18. MANNEN, H. 1975. Reconstruction of axonal trajectory of individual neurons in the spinal cord using Golgi-stained serial sections. J. Comp. Neurol. 159: 357-374. 19. MATSUSHITA, M., M. IKEDA, AND Y. HOSOYA. 1979. The location of spinal neurons with long descending axons (long descending propriospinal tract neurons) in the cat: a study with the horseradish peroxidase technique. J. Camp. Neurol. 184: 63-80. 20. MAUNZ, R. A., N. G. PITTS, AND B. W. PETERSON. 1978. Cat spinoreticular neurons: locations, responses and changes in responses during repetitive stimulation. Bruin Res. 148: 365-379.
21. MEINCK, H.-M., AND B. PIESIUR-STREHLOW. 1981. Reflexes evoked in leg muscles from arm afferents: a propriospinal pathway in man? E.rp. Brain Rex 43: 78-86. 22. MILLER, S., AND J. VAN DER BERG. 1973. The function of long propriospinal pathways in the co-ordination of quadrupedal stepping in the cat. Pages 561-577 in R. B. STEIN, K. G. PEARSON, R. S. SMITH, AND J. B. REDFORD, Eds., Control aJ‘ Posture and Locomotion. (Advances in Behavioral Biology. Vol. 7). Plenum. New York/London. 23. MOLENAAR, I., AND H. G. J. M. KUYPERS. 1978. Cells of origin of propriospinal fibers and of fibers ascending to supraspinal levels. A HRP study in cat and rhesus monkey. Brain
Rex
152: 429-450.
24. PETERSON, B. W., R. A. MAUNZ, N. G. PITTS, AND R. G. MACKEL. 1975. Patterns of projection and branching of reticulospinal neurons. Exp. Brain Res. 23: 333-35 I. 25. REMMEL, R. S., R. D. SKINNER, AND L. B. MINOR. 1980. Eighth nerve activation of cat pontine reticular neurons which project in or near the ascending medial longitudinal fasciculus. Exp. Neurol. 70: 706-7 1I. 26. ROSSI,G. F., AND A. BRODAL. 1957. Terminal distribution of spinoreticular fibers in the cat. Arch. Neurol. Psychiatry (Chicago) 78: 439-453. 27. SCHOMBURG, E. D.. H.-M. MEINCK, J. HAUSTEIN, AND J. ROESLER. 1978. Functional organization of the spinal reflex pathways from forelimb afferents to hindlimb motorneurones in the cat. Brain Res. 139: 2 l-33. 28. SHERRINGTON,C. S.. AND E. E. LASLETT. 1903. Observations on some spinal reflexes and the interconnection of spinal segments. J. Physiol. (London) 29: 58-96. 29. SKINNER, R. D., R. J. ADAMS, AND R. S. REMMEL. 1980. Responses of long descending propriospinal neurons to natural and electrical types of stimuli in cat. Brain Res. 196: 387-403.
472
SKINNER
ET AL.
30. SKINNER, R. D., J. D. COULTER, R. J. ADAMS, AND R. S. REMMEL. 1979. Cells of origin of long descending propriospinal fibers connecting the spinal enlargements in cat and monkey determined by horseradish peroxidase and electrophysiological techniques. J. Comp.
Neural
188: 443-454.
31. SKINNER, R. D., AND R. S. REMMEL. 1978. Monosynaptic inputs to lumbar intemeurons from the lateral vestibulospinal tract and the medial longitudinal fasciculus. Neurosci. Lett. 10: 259-264.
32. SKINNER, R. D., AND R. S. REMMEL. 1980. Monosynaptic supraspinal inputs to long descending propriospinal neurons in cats. Sot. Neurosci. Abstr. 6: 24. 33. THOMAS, R. C., AND V. J. WILSON, 1965. Precise localization of Renshaw cells with a new marking technique. Nature (London) 206: 2 1 l-2 13. 34. WILSON, V. J., AND M. YOSHIDA. 1969. Comparison of effects of stimulation of Deiters’ nucleus and medial longitudinal fasciculus on neck, forelimb, and hindlimb motoneurons. J. Neurophysiol.
32: 743-758.
35. VASILENKO, D. A., AND A. I. KOS~YUKOV. 1976. Brain stem and primary afferent projections to the ventromedial group of propriospinal neurones in the cat. Brain Res. 117: 141-146. 36. YEZIERSKI, R. P., J. L. CULBERSON, AND P. B. BROWN. 1980. Cells of origin of propriospinal connections to cat lumbosacral gray as determined with horseradish peroxidase. Exp. Neural. 69: 493-5 12.