Faculty Bibliography

The morphology of single tensor tympani motoneurons was investigated following antidromic identification and intracellular injection of horseradish peroxidase. Eight motoneurons were selected for complete reconstruction and quantitative analysis. The mean size of tensor tympani somata (26.3 +/- 1.8 micron) make this parvocellular cluster of motoneurons below the trigeminal motor nucleus a population of the smallest cranial motoneurons yet described. Axons emerged from either the soma or a primary dendrite. They coursed dorsolaterally frequently through the trigeminal motor nucleus before looping ventrolaterally into the Vth nerve. No collaterals were observed within the brainstem. The 5 primary dendrites of each cell branched heavily and, on average, exhibited 40 terminal branches with an average tree expansion of 1262.5 micron. The dendritic arborization extended far beyond the nuclear boundaries described by the distribution of cell bodies. These data suggest that the overall membrane area for synaptic innervation is large and thus it provides morphological evidence for the hypothesis that tensor tympani motoneurons receive divergent multisensory synaptic input. The latter assumption was supported by morphological and electrophysiological evidence including close the proximity of motoneuronal dendrites to auditory (superior olivary complex) and somatosensory (trigeminal) relay centers. Since no dendrite ever entered the trigeminal motor nucleus proper the tensor motoneuron pool is distinct from the trigeminal not only in terms of soma size, location and function, but also the disposition and expansion of the postsynaptic receptive field. Based on these criteria the tensor tympani motoneuron pool should no longer be regarded as an accessory trigeminal nucleus but be recognized in its own right as the tensor tympani motor nucleus of V.

Recordings of upper eyelid movements in humans, guinea pigs, and rabbits demonstrated that all three species displayed qualitatively similar patterns of eyelid movement. The relation between amplitude, duration, and maximum velocity in rabbits and humans was nearly identical. Guinea pig blinks were faster than those of rabbit and man. Electromyographic (EMG) recordings in humans demonstrated that the orbicularis oculis muscle participated in downward movement of the upper eyelid during blinks and eyelid closure but did not participate actively in the downward lid movement occurring with gaze changes. When looking straight ahead, the estimated stiffness and viscosity of the upper eyelid were 10 g/mm and 0.38 g X s X mm-1 for humans and 1.17 g/mm and 0.062 g X s X mm-1 for rabbits. Upward and abducting rotations of the eye accompanied blinks in rabbits and guinea pigs. Simultaneously, the eyeball retracted (translational movement) into the orbit. These translational and rotational eye movements resulted from contraction of the retractor bulbi muscle and cocontraction of antagonistic extraocular muscles. The data suggested that humans also retracted the eye during voluntary blinks. The retraction produced a rotation of the eye toward a "primary position" rather than a rotation in one specific direction. The relationship between the maximum velocity, duration, and amplitude of the down phase of a blink may be expressed as a single equation, maximum velocity = c X average velocity, where c is a constant. The same relationship, with a similar value for c, also describes saccadic eye movements and rapid skeletal movements. This implies that all three movements employ comparable neural mechanisms.

Postsynaptic potentials were recorded from motoneurons in the facial nucleus in response to stimulation of the vestibular and trigeminal nerves. The motoneurons were identified by antidromic activation from their peripheral axons. Disynaptic excitatory and inhibitory postsynaptic potentials (EPSPs and IPSPs) and mixed EPSP/IPSPs were recorded in response to vestibular nerve stimulation, ranging in latency from 0.9 to 2.1 ms, with most at 1.5 ms. Activity in secondary vestibular axons recorded within the facial nucleus occurred at a latency of 0.7-1.1 ms. The amplitudes of the vestibular postsynaptic potentials were small, generally less than a millivolt, but double shocks produced marked summation. The average time to peak of ipsilateral vestibular EPSPs, 1.1 ms, was faster than that of either ipsilateral IPSPs, 1.6 ms, or contralateral EPSPs, 1.4 ms. The double-spiked vestibular activity was detectable in double-peaked PSPs. Disynaptic EPSPs, ranging in latency from 2.0 to 3.0 ms, were recorded in response to trigeminal nerve stimulation. The average time to peak was 1.3 ms. The multiple-spiked activity of the trigeminal neurons was detectable in multipeaked EPSPs. Inhibitory ipsilateral effects (Vi IPSPs) were recorded twice as often as excitatory ipsilateral effects (Vi EPSPs), being found in 29% versus 15% of the motoneurons. Contralateral effects were found in 13% of the motoneurons studied, and almost all were excitatory. Analysis of synaptic potential shapes suggested that the excitatory and inhibitory vestibular synapses probably contact distal dendrites preferentially, with the excitatory connections being somewhat closer to the soma. The trigeminal inputs probably contact the facial motoneurons more extensively near the soma. Horseradish peroxidase was injected into the facial nucleus, and retrograde uptake by vestibular neurons was studied. The majority of filled vestibular neurons was ipsilateral to the injection site, especially in the medial vestibular nucleus, ventral y group, and supravestibular nucleus. On the contralateral side, filled vestibular cells were found almost exclusively in the medial nucleus. Filled cells were also noted in the trigeminal nucleus, predominantly ipsilaterally at all rostrocaudal levels. We have demonstrated monosynaptic projections to facial motoneurons from both vestibular and trigeminal nuclei. The trigeminal input is likely to be involved in facial reflexes, especially blinking and grimacing. The afferent vestibular population overlaps that going to the oculomotor and cervical motoneurons; these projections may be collaterals of single vestibular neurons.4+.

Flatfish provide a natural model for the study of adaptive changes in the vestibulo-ocular reflex system. During metamorphosis their vestibular and oculomotor coordinate systems undergo a 90 degree relative displacement. As a result, during swimming movements different types of compensatory eye movements are produced before and after metamorphosis by the same vestibular stimulation. Intracellular staining of central nervous connections in the flatfish with horseradish peroxidase revealed that in postmetamorphic fish secondary horizontal semicircular canal neurons contact vertical eye muscle motoneuron pools on both sides of the brain via pathways that are absent in all other vertebrates studied

The numbers and locations of motoneurons to the stapedius and tensor tympani muscles were determined by retrograde transport of horseradish peroxidase. Stapedius motoneurons lay outside the traditionally recognized facial nucleus, in several distinct locations: (1) in the interface between the facial nucleus and the superior olive; (2) in a thin, scattered lamina of somewhat smaller cells spread dorsal to the facial nucleus; and (3) in a cluster located ventromedial to the rostral third of the facial nucleus. Some cells also lay dorsal to the superior olive or scattered in the reticular formation, just medial to the descending loop of the facial nerve. Tensor tympani motoneurons also lay outside the traditionally recognized trigeminal motor nucleus, in an area just ventral to it. Both motoneuron pools were large, producing innervation ratios that establish stapedius and tensor tympani among the most finely innervated muscles yet studied. The degree of intermingling of large and small cells in these pools may explain, in part, why it has been easier to identify slow muscle fibers physiologically in tensor tympani than in stapedius.

The morphology of secondary vertical vestibular neurons was investigated by injection of horseradish peroxidase (HRP) into cells connected to the posterior canal system in rabbits (lateral-eyed animals) and cats (frontal-eyed animals). Vestibular neurons were identified by stimulation with bipolar electrodes implanted into the ampullae of the anterior and posterior (PC) semicircular canals of pigmented rabbits; in the cat, these cells were identified by natural and electrical stimulation. Axons monosynaptically activated by PC stimulation were injected with HRP in the medial longitudinal fasciculus (MLF). These were later reconstructed by light microscopy after the brains had been processed with a DAB-CoCl2 method. In the rabbit the majority of the axons bifurcated after crossing the midline with one branch ascending and the other descending in the MLF. The ascending branches gave rise to collaterals that terminated in both the trochlear nucleus and the inferior rectus subdivision of the oculomotor nucleus. In addition some axons also sent collaterals into the paramedian pontine reticular formation, the periaqueductal grey and the interstitial nucleus of Cajal. The descending branches were followed to the caudal part of the medulla in the MLF and gave rise to collaterals terminating in the vestibular nuclei, the medullary reticular formation, the perihypoglossal nuclei, the abducens nucleus, and the facial nucleus. In another cell type axons crossed the midline without giving off any collaterals and proceeded caudally in the caudal MLF. The synaptic effects of the two types of cells were concluded to be excitatory and inhibitory, respectively. Cell bodies of contralaterally projecting neurons were located in either the medial or ventro-lateral vestibular nuclei. In the cat we observed two neuron classes, with contralaterally projecting axons, whose synaptic effects are presumably excitatory. Their cell somata were located in the medial vestibular nucleus. Termination patterns were similar to both the trochlear and oculomotor nuclei, but neither projected to the abducens nucleus. One class of neurons was almost identical to that found in the rabbit with the main axon bifurcating in the MLF. The second type lacked a descending branch in the MLF. Axon collaterals of the latter type crossed the midline within the oculomotor nucleus after terminating in the inferior rectus subdivision to reach a similar portion of the ipsilateral oculomotor nucleus. Collaterals of these axons also terminated bilaterally in the supraoculomotor region between trochlear and oculomotor nucleus, the interstitial nucleus of Cajal and prerubral loci (including the fields of Forel). In similarity to the rabbit, presumed inhibitory vestibular neurons were found with axons directed caudally in the MLF without brain stem collaterals.(ABSTRACT TRUNCATED AT 400 WORDS)

Abducens motoneurons and internuclear neurons were identified electrophysiologically in anesthesized, paralyzed cats and stained by intracellular injection of horseradish peroxidase. Neurons were reconstructed and surface area of selected cells measured by light microscopy. Surface area of motoneurons and internuclear neuron with similar soma size and shape were roughly comparable. Dendrites of motoneurons were highly tapered and highly branched. By contrast, dendrites of internuclear neurons were less tapered and less branched. Axons of motoneurons had no collaterals within the brainstem. Internuclear axons crossed the midline at the level of their parent somata and ascended in the medial longitudinal fasciculus toward the oculomotor nucleus. Approximately 30% of the internuclear axons branched in the contralateral medial longitudinal fasciculus sending a fine collateral caudal toward the prepositus hypoglossi nucleus. The results suggest that, on the average, structural correlates of injected neurons (i.e., soma-dendritic morphology) can account at least in part for the earlier firing and higher intraburst frequencies of internuclear neurons versus motoneurons during on-direction rapid eye movements in alert cats.

Motoneurons in the cat oculomotor nucleus have been identified electrophysiologically and stained by intracellular injection of horseradish peroxidase. Axon collateral arborizations with preterminal and terminal boutons identified by light microscopy corresponded to synaptic endings observed by electron microscopy. Despite variations in size and shape, synaptic endings showed similar ultra-structural features and established asymmetrical predominantly axodendritic synaptic contacts usually characterized by the presence of subjunctional dense bodies underlying the postsynaptic membrane densification