Gracile, cuneate, and spinal trigeminal projections to inferior olive in rat and monkey.

In the cat, somatosensory nuclei send substantial projections to the inferior olive, where they terminate in a somatotopic fashion. Although the organization of the cat inferior olive has been used to interpret data from other species, published data suggest this organization may not occur universally. The present study investigated whether the inferior olive in albino rats and cynomolgus monkeys receives the same brainstem somatosensory inputs, whether these inputs are organized somatotopically and, if so, how the organization compares with that in the cat. Projections from the gracile, cuneate and spinal trigeminal nuclei were labeled with wheat germ agglutinin conjugated to horseradish peroxidase or with biotinylated dextran. The results were compared with data from cats (Berkley and Hand [1978] J. Comp. Neurol. 180:253-264). In the rat and monkey, the gracile, cuneate and spinal trigeminal nuclei all project to the contralateral inferior olive, where each nucleus has a distinct preferred terminal field. As in the cat, projections to the medial accessory olive and caudal dorsal accessory olive did not terminate in a precisely organized fashion. Projections to the rostral dorsal accessory olive, however, formed a clear somatotopic map. These somatotopic maps differed from those in the cat in that input from the trigeminal nucleus was confined rostrally, so that the caudal end only received input from the gracile and cuneate nuclei. These data indicate that similar organizational principles characterize the somatosensory projections to the inferior olives of the three species. Nevertheless, distinct species differences occur with regard to the details of this organization.

Ibogaine neurotoxicity: a re-evaluation.

Ibogaine is claimed to be an effective treatment for opiate and stimulant addiction. O'Hearn and Molliver, however, showed that ibogaine causes degeneration of cerebellar Purkinje cells in rats. The present study re-examined cerebellar responses to the high doses of ibogaine used by O'Hearn and Molliver (100 mg/kg or 3 x 100 mg/kg) and sought to determine whether a lower dose (40 mg/kg), one effective in reducing morphine and cocaine self-administration, produced similar responses. Purkinje cell degeneration was evaluated with a Fink-Heimer II stain, and enhanced glial cell activity with an antibody to glial fibrillary acidic protein. Every rat treated with the high dose of ibogaine displayed clear evidence of Purkinje cell degeneration. The degeneration consistently occurred in the intermediate and lateral cerebellum, as well as the vermis. Purkinje cells in lobules 5 and 6 were particularly susceptible. Given the response properties of cells in these lobules, this finding suggests any long-term motor deficits produced by ibogaine-induced degeneration should preferentially affect the head and upper extremity. In marked contrast, rats given the smaller dose of ibogaine displayed no degeneration above the level seen in saline-treated animals. When combined with information on other compounds, these data suggest that the degenerative and "anti-addictive' properties of ibogaine reflect different actions of the drug.

18-Methoxycoronaridine, a non-toxic iboga alkaloid congener: effects on morphine and cocaine self-administration and on mesolimbic dopamine release in rats.

Ibogaine, a naturally occurring iboga alkaloid, has been claimed to be effective in treating addiction to opioids and stimulants, and has been reported to inhibit morphine and cocaine self-administration in rats. However, ibogaine also has acute nonspecific side effects (e.g. tremors, decreased motivated behavior in general) as well as neurotoxic effects (Purkinje cell loss) manifested in the vermis of the cerebellum. 18-Methoxycoronaridine (MC) is a novel, synthetic iboga alkaloid congener that mimics ibogaine's effects on drug self-administration without appearing to have ibogaine's other adverse effects. Acutely, in rats, MC decreased morphine and cocaine self-administration but did not affect bar-press responding for water. In some rats, treatment with MC (40 mg/kg) induced prolonged decreases in morphine or cocaine intake lasting several days or weeks. MC had no apparent tremorigenic effect, and there was no evidence of cerebellar toxicity after a high dose (100 mg/kg) of MC. Similar to the effects of ibogaine and other iboga alkaloids that inhibit drug self-administration, MC (40 mg/kg) decreased extracellular levels of dopamine in the nucleus accumbens. MC therefore appears to be a safer, ibogaine-like agent that might be useful in the treatment of addictive disorders.

Gracile projection to the cat medial accessory olive: ultrastructural termination patterns and convergence with spino-olivary projection.

The caudal medial accessory subdivision of the inferior olive (cMAO) receives information from the hindlimb from both the gracile nucleus and the lumbosacral spinal cord. This study determined which elements in cMAO serve as the postsynaptic targets of the gracile projection and whether these elements also receive input from the lumbosacral spinal cord. Gracile axons were labeled in cats by anterograde transport of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP), visualized with tetramethylbenzidine. Convergence of gracile and lumbosacral axons was evaluated by labeling in the same animal, one pathway by WGA-HRP and the other by degeneration. In cMAO, gracile axons synapse with equal probability on dendritic spines and distal dendritic shafts. This termination pattern contrasts markedly with that of other somatosensory inputs to the inferior olive and may account for the greater heterogeneity in responses to somatosensory stimuli displayed by neurons in cMAO. The distal dendritic shafts receiving gracile input were more likely than dendritic spines to receive convergent input from putative inhibitory synapses. The most likely source of these inhibitory synapses is the parasolitary nucleus, a structure that has been shown by others to receive input from the cerebellum. Thus the parasolitary nucleus may serve as an inhibitory relay between the cerebellum and cMAO. The dendritic spines in cMAO that receive input from the gracile nucleus often receive additional input from the lumbosacral spinal cord. This convergence of somatosensory axons on dendritic spines may provide a mechanism through which the unusually complex receptive fields of neurons in cMAO are generated.

Functional properties of dorsal horn neurons that project to the dorsal accessory olive.

1. This study examined the responses to natural cutaneous stimuli of neurons in the dorsal horn of the lumbosacral spinal cord that project to the dorsal accessory portion of the inferior olive (DAO) in cats anesthetized with pentobarbital sodium. Extracellular activity was recorded from single units antidromically activated by currents of less than or equal to 70 muA applied to DAO. 2. A total of 119 antidromically activated neurons was examined. Their antidromic activation latencies displayed a wide range (2.5-24.6 ms). The average latency corresponds to a conduction velocity of 24 m/s. 3. Collision was demonstrated for 24 neurons. All responded to some form of natural cutaneous stimulation. Their receptive fields encompassed some portion of the hind limb, particularly the toes; one-third displayed gradients of sensitivity. 4. Based on their thresholds to peripheral stimulation, the 24 neurons fell into five categories, those sensitive to light cutaneous stimuli (i.e., hair movement or light touch; 37.5%), rub (21%), tap (21%), pressure (12.5%), or noxious stimuli (8%). 5. Comparison of these results with data on the other major source of somatosensory information for DAO, the gracile nucleus (examined previously with the same methods), suggests that the sensitivity of neurons in DAO to light cutaneous stimuli is mediated primarily by neurons in the dorsal horn. The sensitivity of neurons in DAO to tap, rub, or pressure, on the other hand, might be mediated by neurons in either the dorsal horn, the gracile nucleus, or both.

Spino-olivary termination on spines in cat medial accessory olive.

Three functional regions of the inferior olive, the caudal medial accessory olive (cMAO) and the caudal and rostral dorsal accessory olive (DAO) receive input from the spinal cord. The present study determined how spinal inputs to cMAO interact with olivary neurons. These inputs were labeled by injections in cat lumbosacral of wheat germ agglutinin conjugated to horseradish peroxidase. The tracer was visualized with tetramethylbenzidine. The morphology of the labeled spino-olivary terminals and the relationship between these terminals and postsynaptic elements were determined. Spino-olivary terminals in cMAO displayed the morphological characteristics classically associated with excitatory synapses. Almost three quarters synapsed on spines, most of which contacted other spines, forming spine clusters. The majority of postsynaptic spines also received convergent input from apparently excitatory, nonlumbosacral afferents. This postsynaptic organization provides several possible benefits for the putative role of cMAO in the control of posture. An earlier study demonstrated that in DAO, almost three quarters of lumbosacral, spino-olivary terminals synapse on dendrites (Molinari: Neuroscience 27:425-435, 1988). Thus, lumbosacral afferents appear to differ fundamentally in the way in which they interact with neurons in cMAO and DAO. These results suggest that the way information is processed may be as important in determining the functional differences between olivary regions as what information is received.

Variations in the cellular incorporation of [3H]proline.

Injections of [3H]proline into various CNS nuclei in the cat and into the rat dorsal column nuclei labeled terminal targets of those regions despite the fact that only glia (and not neuronal somata) were labeled at the injection site. In contrast, although injections into cat CNS fiber tracts labeled glial cells surrounding the axons, neither the axoplasm, somata or terminals of those axons were labeled. In addition, injections into cat PNS dorsal root ganglia labeled ganglion cells. These results support previous findings suggesting that 3H-proline is incorporated into a class of molecules in the CNS (but not PNS) that is transferred from glia into neuronal somata (but not axoplasm).

Ultrastructural heterogeneity of spinal terminations in the cat inferior olive.

Three different functional regions of the inferior olive receive direct input from the spinal cord. The present study examined spinal termination patterns in two of these functional entities, the rostral and caudal halves of the dorsal accessory olive, with anterograde transport of wheat germ agglutinin conjugated with horseradish peroxidase visualized by tetramethylbenzidine. The injections primarily included the spino-olivary projection from neurons in the lumbosacral dorsal horn. Two types of labeled terminals were found, small end bulbs (84%) and large en passant boutons (16%). The small end bulbs displayed distinct rostrocaudal variations in their termination patterns. In the rostral dorsal accessory olive they synapsed most frequently on dendrites that directly contacted other dendrites, forming dendritic thickets. In the caudal dorsal accessory olive, they synapsed less often in thickets and more often on isolated dendrites. Conversely, the large, en passant boutons synapsed primarily in thickets and failed to display comparable rostrocaudal shifts. All somatic afferents in the rostral dorsal accessory olive examined to date synapse primarily in dendritic thickets, suggesting that the thicket is a major site through which neurons in that region detect peripheral somatic events. Within dendritic thickets, single afferents often contact multiple dendrites and the dendrites, in turn, sometimes give rise to spinous processes. It is proposed that these spinous processes participate in synaptic glomeruli, which others have shown to be the primary targets of cerebellar afferents to the dorsal accessory olive. These results suggest that somatic afferents to the rostral dorsal accessory olive influence a greater number of neurons and are more likely to interact with cerebellar input than are somatic afferents to the caudal region. This possibility is consistent with the more complex types of movement influenced by the rostral compared with the caudal halves of the dorsal accessory olive.

Ultrastructure of the gracile nucleus projection to the dorsal accessory subdivision of the cat inferior olive.

This study examined the termination pattern within the dorsal accessory subdivision of the cat inferior olive of axons arising from the gracile nucleus. The gracile terminals were labeled by anterograde transport of wheat germ agglutinin complexed to horseradish peroxidase and visualized with tetramethyl benzidine. Gracile terminals were found to contain round synaptic vesicles and form asymmetric synaptic contacts. Of particular interest was the finding that gracile axons, like axons from the spinal cord, terminate primarily outside of synaptic glomeruli. Yet most of the gracile terminals did not synapse on isolated dendritic elements. Rather, the majority contacted distal dendrites which directly contacted other dendritic elements, forming simple complexes termed dendritic thickets. Typically the dendritic thickets were composed of two or three dendrites that received input from more than one round vesicle-containing synaptic terminal. Only one terminal per thicket was labeled by injections in the gracile nucleus. This clustering of pre- and postsynaptic elements within the thickets provides opportunities for many of the same interactions allowed by synaptic glomeruli, in particular divergence and convergence of information.

Functional properties of neurons in the cat gracile nucleus that project to the dorsal accessory olive.

This study evaluated the functional properties of neurons in the gracile nucleus that project to the dorsal accessory portion of the inferior olive (DAO) and compared these with properties of other efferents from the dorsal column nuclei (DCN). Projection neurons were identified in anesthetized cats by microstimulation within DAO. They were further tested to insure that they could not be antidromically activated by stimulation in the medial lemniscus. Forty percent of the DAO projection neurons failed to respond to any form of natural stimulation. Of the 60% that did, most: 1) required stimuli of intensities greater than those needed to activate peripheral mechanoreceptors but less than those needed to excite peripheral nociceptors; 2) had receptive fields on the contralateral hindlimb or tail; and 3) could not follow peripheral stimuli presented at rates greater than 20 Hz. The conduction velocities of DAO projection neurons averaged only 4 m/s. In contrast, most of the neurons with axons in the medial lemniscus responded to very light cutaneous stimuli and followed peripheral stimuli presented at rates up to 100 Hz. Their average conduction velocity was 14 m/s. The low conduction velocity of the DAO projection neurons is unusual for DCN but is consistent with published descriptions of DAO. The limited responsiveness of the DAO projection neurons is, however, not uncharacteristic of DCN but places this projection at one end of the spectrum of DCN efferents, with the diencephalic projection at the other end. It is proposed that the DCN provide a filtered (and possibly movement-modulated) signal to the inferior olive for use in the control of movement.

Ascending somatosensory projections to the medial accessory portion of the inferior olive: a retrograde study in cats.

The cells in the dorsal column nuclei, lumbosacral spinal cord, lateral cervical nucleus, and nucleus z that project to the medial accessory portion of the inferior olive of cats were identified with retrograde tracing techniques. Injections of wheat germ agglutinin complexed to horseradish peroxidase were made in the caudal portion of the medial accessory olive, either (1) involving no portion of the dorsal accessory olive or (2) involving in addition the caudal tip of the dorsal accessory olive. The tissue was processed with tetramethyl benzidine. The locations of all relay neurons were compared with those of dorsal accessory olive projection neurons, as described in a previous study (Molinari, '84a). Localized populations of neurons gave rise to most of the projection to the medial accessory olive. These neurons were found in the peripheral portions of the dorsal column nuclei caudal to the obex and in the ventromedial ventral horn of the entire lumbosacral enlargement. Few projection neurons were found in the lateral cervical nucleus and none in the nucleus z. Neurons in the peripheral dorsal column nuclei and ventromedial ventral horn were labeled by injections in either the medial or dorsal accessory olives. Following medial accessory olive injections, however, they constituted the only labeled somatosensory neurons, while following dorsal accessory olive injections they represented only a small fraction of the labeled neurons. Based on their locations, it is proposed that these neurons might be the source for both the medial and dorsal accessory olives of information signalling movement of the proximal limb. Such a proposal is consistent with functional descriptions of the medial and dorsal accessory olives and the cerebellar anterior lobe.

The primate as a model for the human temperature-sensing system: 1. Adapting temperature and intensity of thermal stimuli.

The thermal sensitivity of three humans and two rhesus monkeys was measured behaviorally, using the "yes-no" paradigm of the Theory of Signal Detection. The aim was to evaluate the monkey's thermal-sensing system as a model for that of humans. Three of the principal variables of human thermal sensations--rate of the temperature change, area of stimulation, and site of stimulation--were held constant. The other three variables--adapting skin temperature (AT), intensity, and direction of the temperature change--were varied systematically. Systematic differences between species were not evident for warming or cooling stimuli. Isodetectability curves (d'e = 1) for small cooling stimuli plotted as a function of the AT were isomorphic, and the points for the human and monkey subjects were frequently superimposed. Isodetectability curves for warming stimuli, on the other hand, had similar shapes for ATs between 33 degrees and 40 degrees C, but the points for the different subjects were not superimposed. At ATs below 30 degrees C, one of the two humans in the warming series and the two monkeys continued to show similarly shaped curves, but the other human was markedly different. Qualitative descriptions of the thermal sensations obtained during threshold measurements of human subjects, reported previously, suggest that this unusual subject probably adopted a criterion qualitatively different from that used by the other subjects. The data presented here and in combination with previously published work from this laboratory (Kenshalo, 1970) suggest that thermal stimuli produce similar sensations in rhesus monkeys and humans, and that the neural systems responsible for coding AT and temperature change in the two species are fundamentally similar.

Ascending somatosensory projections to the dorsal accessory olive: an anatomical study in cats.

The cells in the gracile nucleus that project to the dorsal accessory olive were identified in cats with retrograde tracing techniques. In the same animals, the retrograde labeling patterns in the lateral cervical nucleus and the lumbosacral spinal cord were also examined. Small injections of wheat germ agglutinin complexed to horseradish peroxidase were made in the ventrolateral portion of the dorsal accessory olive without involving the medial accessory olive and without damaging the medial lemniscus. The tissue was processed with tetramethyl benzidine. In each of the three relay nuclei, the neurons that project to the ventrolateral portion of the contralateral dorsal accessory olive are highly concentrated in small regions. In the gracile nucleus these cells are found almost exclusively in the transitional zone, just caudal to the obex and rostral to the clusters. In the lateral cervical nucleus they are concentrated in the dorsolateral tip. In the lumbosacral spinal cord in segments L5 and L6 the cells are found primarily in lamina V, while in segments L7 and S1 they are found along the ventromedial edge of the ventral horn. In the gracile and lateral cervical nuclei there is no segregation of neurons that project to the rostral and caudal portions of the dorsal accessory olive. Comparison of these results with physiological data suggests that each of the three sources of ascending somatic input conveys some distinct aspect of sensory information from the hindlimb to the dorsal accessory olive.

The cutaneous sensitivity of units in laminae VII and VIII of the cat.

The cutaneous sensitivity of units in laminae VII and VIII was compared to that of dorsal horn units in the same cats. There appeared to be a reversal of somatotopic organization between the dorsal and ventral horns. When this reversal was taken into account, it was apparent that most ventral horn units have localized receptive fields. They responded almost exclusively to intense, presumably noxious, stimulation of these discrete skin regions.