Search

Electron microscopic studies of the marginal zone of the dorsal horn in the cat spinal cord revealed 4 different types of synaptic endings contacting marginal neurons. These included boutons with clear spherical vesicles, boutons with clear flattened vesicles, boutons with a mixture of clear spherical and flattened vesicles and boutons with a mixture of clear spherical and dense-cored vesicles. The origin of the synaptic input to marginal neurons was determined by means of degeneration studies at the ultrastructural level following various surgical lesions. It was found that primary afferents contribute approximately 40% of the boutons on the distal dendrites, 20% on the proximal dendrites and 10% on the soma of marginal neurons, while the axons of the tract of Lissauer, which stem from gelatinosa neurons and are inhibitory in function, contribute approximately 20% of the boutons on the distal dendrites, 60% of those on the proximal dendrites and 50% of the somatic endings on marginal neurons. Cranial nerve afferents contribute approximately 10% of the boutons on each area of the marginal neurons within the first cervical segment; this converging afferent input suggests that these cells function as non-specific central nociceptive relay neurons. Fluorescence histochemical studies for the demonstration of biogenic amines revealed a fairly dense array of fluorescent fibers and terminals in the marginal zone. There was no change in this fluorescence following the surgical lesions and their origin remains unknown. Although the observed synaptology of the marginal zone is more complex than that upon which the central inhibitory balance theory of pain was originally proposed, our data provide good anatomic support for that theory.

Cortical recruiting activated from the midline thalamus was studied in 80 animals, including (1) distribution elicited at different stimulus positions in the AP and transverse planes, (2) comparative latencies from parasagittal points to the cortices of the two hemispheres, and (3) the effects of acute and pf chronic lesions involving conceivable sources and routes of activation. The following are our significant conclusions: 1. 1. Using stimulus trains (with stimuli 5 msec. apart) to replace usual single stimuli, the negative phase of the single response evoked from a recruiting region is progressively extended as the length of a train increases. When trains at 6 per sec. recurrence replace the usual recruiting stimulus, the wave durations are also prolonged, although with longer trains the response ordinarily falls off rapidly past 25 stimuli. 2. 2. From parasagittal stimulation there is a significant latency difference to ipsi- and contralateral recording points, only midline stimulation resulting in nearly equal latency. Parasagittal placement shortens latency to the ipsilateral and lengthens it to the contralateral cortex. 3. 3. Neither acute nor chronic lesions of the corpus callosum influence recruiting activated from the midline thalamus. Lesions of the anterior limb of the internal capsule reduce the responsiveness of the rostral cortex of that hemisphere significantly. If a large area of retrograde change is produced by a chronic lesion, the effect upon recruiting is more pronounced than with a smaller area. Parasagittal lesions which do not affect the internal capsule or penetrate the thalamic peduncle interfere with recruiting much less significantly. Combinations of the two nearly abolish ipsilateral recruiting. 4. 4. Very active recruiting can be obtained from midline nuclei after destruction of both dorsomedial nuclei. Lesions on the raphe involving centralis medialis and rhomboideus nuclei significantly affect recruiting from electrode penetrations in the neighborhood of the lesion. Due to destruction of commissural connections the effect is more pronounced contralaterally. Active ipsilateral recruiting can still be elicited from rostral stimulating positions in the midline thalamus when a large lesion has destroyed the centralis medialis. Small dorsal lesions of the raphe reduce contralateral recruiting and elevate its threshold, indicating that the commissural system concerned with recruiting does not occupy a restricted part of the raphe. 5. 5. Section of the anterior limb of the internal capsule effectively abolishes the repetitive cortical spindles that follow single stimuli to the intralaminar nuclei. 6. 6. The findings are in general accord with prevailing opinion, give direct confirmation of the importance of the midline nuclei, and show that the dorsomedial nuclei, at least, are not essential to recruiting.

Three pupillodilator pathways have been identified in the brain stem of the cat by means of electrical stimulation with microelectrodes. 1. (1) A descending, excitatory pathway which projects to the intermediolateral cell column (C8-T2) has been identified in the ventrolateral tegmentum. This pathway has been traced at successively caudal levels. It lies dorsolateral to the medial lemniscus at the collicular and pontine levels and next to the spinothalamic tract. At more caudal levels, the pathway moves to a position dorsolateral to the pyramidal tract. In the cervical spinal cord, it shifts to a superficial position in the lateral funiculus near the insertion zone of the dentate ligament. Axon degeneration studies confirm these results and indicate that very few fibers from the tegmentum reach the spinal cord directly. This pathway projects primarily in an ipsilateral manner. Some of the axons cross in the dorsal commissure of the spinal cord at the C8-T2 levels to reach the contralateral side. 2. (2) An ascending inhibitory pathway projecting to the Edinger-Westphal nucleus has been identified in the paramedian reticular formation of the medulla and pons. This pathway has been traced rostrally through the medial tegmental reticular formation in an area dorsal to the medial lemniscus. At rostral levels, it lies lateral to the interpeduncular nucleus. 3. (3) An ascending inhibitory pathway has been found to be associated with the lateral pupillodilator pathway. It seems probable that this system is involved in nociceptive reflex pupillary dilatation. Whether this pathway is part of the spinothalamic tract or is a separate pathway has not been settled.

The plasticity of cervical primary afferents has been investigated in young adult cats in which the trigeminal root, together with most of the ganglion, had been excised via a new approach to these structures during the first week after birth. Once degeneration debris had disappeared, bilateral ganglionectomies of the upper three cervical dorsal roots were done in five animals and the IXth and Xth roots transected on the side of the chronic trigeminal denervation in one instance; the degeneration pattern on the chronically denervated side was compared to that on the normal side. There was mild evidence of increased degeneration of the cervical afferents in the C1 segment and in the medulla (subnucleus interpolaris of V) on the chronically denervated side. The proliferation was more apparent in the kitten operated at three days of age and was only vestigial in animals operated at six or seven days; it had no appreciable tendency to extend into the large contiguous pool of denervated trigeminal neurons. There was no evidence of sprouting of the IXth and Xth nerve afferents. These observations indicate that primary afferents of the upper cervical roots and of the glossopharyngeal and vagus nerves have little neuroplastic potential even at early stages. The literature on neuroplasticity is controversial and a brief review of this together with suggestions regarding the reasons for some of these conflicts is presented. Some structural and functional effects of chronic trigeminal denervation are briefly described.

The ascending degeneration resulting from experimental lesions of the ventral funiculus of the spinal cord of Macaca mulatta has been studied using the Nauta technique and its variants. The ventral spinothalamic tract is shown to be an independent entity with respect to the lateral spinothalamic tract; its fibers are widely distributed in the ventral funiculus and it establishes connections with the brain stem and thalamus which are analogous but not identical to those of the latter. Its role in the relay of nociceptive input is discussed in view of the similarity in hodology of the two systems and it is proposed that it may be responsible for the failure of anterolateral cordotomy to control pain on a long term basis. Other ascending systems in the ventral funiculus include the spino-olivary and spino-reticular tracts, as well as minor connections to the N. of Edinger-Westphal, the red nucleus and the superior colliculus. The projections from the ventral quadrant of the spinal cord to the brain stem are almost entirely ipsilateral until the rostral mesencephalon is reached, at which level the N. of Darkschewitz receives both ipsilateral and crossed input; the magnocellular nucleus of the medial geniculate body receives a small contribution which is mainly ipsilateral. In the thalamus the VPL receives predominantly ipsilateral projections while the input to the paralaminar nuclei is only slightly less pronounced contralaterally than ipsilaterally.

The ascending projections of the dorsolateral funiculus of the spinal cord to the brain stem have been determined in five macaque monkeys. Connections to the lateral cervical nucleus and to the reticular nucleus of the cord in the C1 and C2 segments are present. In the medulla the most prominent connections are to the nuclei "Z" and "x" of Brodal and Pompeiano, to the rostral portion of n. gracilis and to the n. proprius of the restiform body. Minor projections reach the rostral part of the medial and lateral cuneate nuclei, the reticular nucleus, the n. centralis dorsalis and the periependymal gray. There were no projections to planes rostral to the medulla. In view of the connections established it is concluded that ascending systems in the DLF to the brain stem of primates are concerned with transmission of mechanoreceptor input to the cerebellum and thalamus and that nociceptive relay appears very unlikely.

The possibility that collateral sprouting might occur in the convergence area of trigeminal and cervical afferents in the dorsal horn of the C1 and C2 segment was evaluated by electrophysiologic methods. Two groups of cats were studied. In one group the trigeminal root was sectioned unilaterally and, after survival periods in excess of 6 months, they were decerebrated and and the characteristics and distribution of dorsal cord potentials evoked by stimulating simultaneously both greater auricular nerves were studied. In the second group of animals, dorsal root ganglionectomies of C1, C2, and C3 were carried out and, after similar survival periods, the dorsal cord potentials evoked by simultaneous stimulation of both infraorbital nerves were examined and compared. In both groups of animals, microelectrode studies were carried out on units in the areas of the denervated nucleus from which the dorsal cord potentials had been obtained. Comparison of external recordings from the control and denervated sides demonstrated no differences in the configuration or distribution of the evoked responses in either group. In the single unit studies no evidence of neuronal burst firing or spontaneous hyperactivity was encountered. These observations indicate (i) that neuroplasticity of primary afferents in this area is either absent or of too limited degree to be detectable by the methods employed and (ii) that after 6 months of primary denervation, interneurons no longer show the spontaneous hyperactivity which has been reported to occur after shorter intervals.

Summary The organization of primary afferents from the cranial and upper cervical root systems in the subnucleus caudalis and uppermost cervical dorsal horn has been investigated in cats. Following transection of a single or multiple sensory roots the distribution of primary fibers in the spinal tract and tract of Lissauer and of terminal knobs in the nucleus was studied by light and electron microscopic methods. A considerable proportion of the small myelinated fibers and most of the unmyelinated fibers in the spinal tract of V are not of dorsal root origin. Fibers of the trigeminal root reach the lower half of the C 2 segment in minimal numbers. The marginal zone receives primary axodendritic contacts on small dendrites and none of the knobs in the characteristic clusters are of primary origin. Marginal neuron somata do not appear to receive primary afferents. The marginal plexus has a large primary contingent and it is concluded that it serves as system for redistributing fibers in the tract on their way to the nucleus. The substantia gelatinosa contains numerous primary endings which are distributed in a patchy manner; the majority establish axodendritic synapses. A large proportion of these primary endings are postsynaptic to knobs which contain flattened vesicles and rarely spherical vesicles. Occasional primary endings are presynaptic to normal knobs which contain spherical vesicles. Primary endings on gelatinosa neurons are present though infrequent. In the nucleus proprius, small primary knobs synapse on medium and small caliber dendrites but not on neuron somata. Large endings showing pleomorphism of synaptic vesicles which may be indicative of a degenerative process were observed in synaptic contact with small and medium dendrites and with cell bodies. Both types of endings were seen to be postsynaptic to small knobs which contained either spherical or flattened vesicles. The functional significance of these obsevations is discussed.

The structural relationships between primary trigeminal and cervical dorsal root afferents in the spinal cord and the medulla were investigated in 13 cats. In 5 cats rhizotomies of one upper cervical dorsal root and of the contralateral trigeminal dorsal root were done at one operative session. In the remaining 8 cats rhizotomies of from 1 to 3 cervical roots were performed. The resultant degeneration was studied by the Nauta-Gygax and Fink-Heimer methods. The degree of anatomical overlap between afferents of the first 3 upper cervical roots and trigeminal terminals in the dorsal horn at identical levels is maximal in the lower half of C1 and upper third of C2. Minimal overlap of C4 with trigeminal endings occurs, while no afferents from the C5 to C8 roots reach the caudalmost endings of the trigeminal. Convergence of cervical with trigeminal afferents is also seen in the spinal nucleus of V in the low medulla in two discrete loci.

Because of the intimate structural and functional relationships between cervical and trigeminal primary afferents in the caudalmost portion of the subnucleus caudalis of the trigeminal nerve, the possibility that cervical afferents might sprout to occupy vacated synaptic sites when the trigeminal afferents degenerated was investigated. Four adult cats were kept for periods of 11.5–24 months after trigeminal rhizotomy at which time symmetrical upper cervical spinal ganglionectomies were done. The ensuing degeneration was studied by the Nauta-Gygax and Fink-Heimer I methods at 6–13 days postoperatively. No evidence could be found that sprouting of cervical primary afferents had occurred at any level on the chronically denervated side. The significance of these findings is discussed in the light of past and recent observations on the problem of plasticity of connections in the mammalian central nervous system.

The dorsal motor nucleus of the vagus was destroyed in cats. After periods of survival ranging from 9 to 44 days the response of heart, bronchi, esophagus, and duodenum to stimulation of the distal end of the cut cervical vagus was within the normal range or slightly depressed. It is concluded that the dorsal motor nucleus is not a major source of visceromotor fibers.

Summary The spinal tract and subnucleus caudalis of the trigeminal nerve of the cat have been studied with light and electron microscopic techniques. The changing pattern of the spinal tract in its descent is described. The marginal zone, in addition to the typical large marginal neurons, is characterized by a prominent tangential plexus and an extremely profuse population of end knobs, the great majority of which contain spherical vesicles; these axonal terminals are arranged in conspicuous clusters in which knobs are piled densely one on another. Some of them establish axo-axonic contacts in addition to axodendritic synapses. This layer also contains the greatest number of knobs with dense core vesicles and these establish axodendritic as well as axosomatic contacts with marginal neurons. Large dendrites with prominent spines are present both in the marginal layer and in the nucleus proprius. The substantia gelatinosa contains small neurons which usually have few axosomatic contacts; these are formed by small knobs which usually contain spherical vesicles and the synapses are predominantly of the symmetrical type. In the neuropil of the substantia gelatinosa the great majority of the axon terminals are of small size; most of these contain spherical vesicles with occasional prominent dense core granules; the contacts are mostly axodendritic and asymmetrical, and moderate numbers of axo-axonic contacts and glomeruli are present. The nucleus proprius is relatively homogenous throughout, contains large and small neurons each variety having either many or few axosomatic contacts by knobs containing either spherical or flat vesicles. Axodendritic synapses are the most common and large axon terminals which often establish multisynaptic contacts are characteristically present in this part of the nucleus and occasionally in the marginal zone.