Motoneuron axon distribution in the cat stapedius muscle.

Stapedius-motoneuron cell bodies in the brainstem are spatially organized according to their acoustic response laterality, as demonstrated by intracellular labeling of physiologically identified motoneurons [Vacher et al., 1989. J. Comp. Neurol. 289, 401-415]. To determine whether a similar functional spatial segregation is present in the muscle, we traced physiologically identified, labeled axons into the stapedius muscle. Ten labeled axons were visible in the facial nerve and five could be traced to endplates within the muscle. These five axons had 39 observed branches (others may have been missed). This indicates an average innervation ratio (> or = 7.8) which is much higher than that obtained from previous estimates of the numbers of stapedius motoneurons and muscle fibers in the cat. One well-labeled stapedius motor axon innervated only a single muscle fiber. In contrast, two labeled axons had over 10 endings and innervated muscle fibers spread over wide areas in the muscle. Two of the axons branched and coursed through two primary stapedius fascicles, indicating that the muscle zones innervated by different primary fascicles are not functionally segregated. In another series of experiments, retrograde tracers were deposited in individual primary nerve fascicles. In every case, labeled stapedius-motoneuron cell bodies were found in each of the physiologically identified stapedius-motoneuron regions in the brainstem. These observations suggest there is little, if any, functional spatial segregation based on separate muscle compartments in the stapedius muscle, despite there being functional spatial segregation in the stapedius-motoneuron pool centrally.

Generators of the brainstem auditory evoked potential in cat. I. An experimental approach to their identification.

This paper is the first in a series aimed at identifying the cellular generators of the brainstem auditory evoked potential (BAEP) in cats. The approach involves (1) developing experimental procedures for making small selective lesions and determining the corresponding changes in BAEP waveforms, (2) identifying brainstem regions involved in BAEP generation by examining the effects of lesions on the BAEP and (3) identifying specific cell populations involved by combining the lesion results with electrophysiological and anatomical information from other kinds of studies. We created lesions in the lower brainstem by injecting kainic acid which is generally toxic for neuronal cell bodies but not for axons and terminals. This first paper describes the justifications for using kainic acid, explains the associated problems, and develops a methodology that addresses the main difficulties. The issues and aspects of the specific methods are generally applicable to physiological and anatomical studies using any neurotoxin, as well as to the present BAEP study. The methods chosen involved (1) measuring the BAEP at regular intervals until it reached a post-injection steady state and perfusing the animals with fixative shortly after the last BAEP recordings were made, (2) using objective criteria to distinguish injection-related BAEP changes from unrelated ones, (3) making control injections to identify effects not due to kainic acid toxicity, (4) verifying the anatomical and functional integrity of axons in lesioned regions, and (5) examining injected brainstems microscopically for cell loss and cellular abnormalities indicating dysfunction. This combination of methods enabled us to identify BAEP changes which are clearly correlated with lesion locations.

Acoustic reflex frequency selectivity in single stapedius motoneurons of the cat.

1. The sound frequency selectivities of single stapedius motoneurons were investigated in ketamine anesthetized and in decerebrate cats by recording from axons in the small nerve fascicles entering the stapedius muscle. 2. Stapedius motoneuron tuning curves (TCs) were very broad, similar to the tuning of the overall acoustic reflexes as determined by electromyographic recordings. The lowest thresholds were usually for sound frequencies between 1 and 2 kHz, although many TCs also had a second sensitive region in the 6- to 12-kHz range. The broad tuning of stapedius motoneurons implies that inputs derived from different cochlear frequency regions (which are narrowly tuned) must converge at a point central to the stapedius motoneuron outputs, possibly at the motoneuron somata. 3. There were only small differences in tuning among the four previously described groups of stapedius motoneurons categorized by sensitivity to ipsilateral and contralateral sound. The gradation in high-frequency versus low-frequency sensitivity across motoneurons suggests there are not distinct subgroups of stapedius motoneurons, based on their TCs. 4. The thresholds and shapes of stapedius motoneuron TCs support the hypothesis that the stapedius acoustic reflex is triggered by summed activity of low-spontaneous-rate auditory nerve fibers with both low and high characteristic frequencies (CFs). Excitation of high-CF auditory nerve fibers by sound in their TC "tails" is probably an important factor in eliciting the reflex. 5. In general, the most sensitive frequency for stapedius motoneurons is higher than the frequency at which stapedius contractions produce the greatest attenuation of middle ear transmission. We argue that this is true because the main function of the stapedius acoustic reflex is to reduce the masking of responses to high-frequency sounds produced by low-frequency sounds.

Brainstem facial-motor pathways from two distinct groups of stapedius motoneurons in the cat.

To determine the brainstem origins and axonal routes of stapedius motoneurons, we labeled motoneurons by injecting cat stapedius muscles with horseradish peroxidase. Some injections were made in normal cats and some in cats in which the middle segment of the internal facial genu had been cut. By tracing labeled axons and by comparing the locations of labeled cell bodies in normal and lesioned cats, we divided stapedius motoneurons into two groups: "perifacial" and "accessory." Perifacial stapedius motoneurons have cell bodies located around the motor nucleus of the facial nerve and axons which follow the classical course of facial motor axons through the internal genu of the facial nerve. Accessory stapedius motoneurons have cell bodies near the descending facial motor root and axons which ascend to the rostral tip of the internal facial genu, abruptly reverse direction, and then join the descending facial motor root. The sharply hooked course of axons of accessory stapedius motoneurons is similar to the course of axons from other accessory nuclei of cranial nerves V-VII. Our present results, with those of McCue and Guinan (J. Neurophysiol. 60:1160-1180, '88), demonstrate that cats have two groups of stapedius motoneurons which can be separated anatomically by the locations of their cell bodies or by the courses of their axons, and which, on the average, have different response properties.

Number and distribution of stapedius motoneurons in cats.

Cell bodies of stapedius motoneurons were identified by retrograde transport of horseradish peroxidase (HRP) following injections into the stapedius muscle. Large injections were made in an attempt to label all stapedius motoneurons. To control for labeling of non-stapedial neurons resulting from spread of HRP, we determined the locations of brainstem neurons labeled by HRP applied to the facial nerve, the chorda tympani nerve, the auricular branch of the vagus nerve, the tensor tympani muscle, and the cochlea. In three cats analyzed in detail, 1,133-1,178 neurons projecting to the stapedius muscle were identified. Arguments are given which suggest that in these three cats all stapedius motoneurons were labeled. The labeled stapedius neurons may all be motoneurons because they all stain positively for acetylcholinesterase and have medium-coarse Nissl bodies. Most stapedius motoneurons were located around the motor nucleus of the facial nerve. Staphedius motoneurons were also found near the descending limb of the facial-nerve root, in the peri-olivary neuropil, and in the reticular formation with the ascending fibers of the facial-nerve root.

Topographic organization of the olivocochlear projections from the lateral and medial zones of the superior olivary complex.

By anterograde tracing using autoradiography, we have found topographic organizations in the projections of both medial and lateral olivocochlear (OC) systems in the cat. Lateral-zone injections show an ipsilateral cochleotopic projection pattern with more medial injections projecting more basally in the cochlea. In the contralateral cochlea, in contrast, the projections from all of the lateral-zone injections were predominantly to the apex. However, detailed analysis suggests the possibility that the contralateral lateral-zone projections may have the same cochleotopic organization as the ipsilateral projections but with a heavy bias toward the apex. Medial-zone injections show a pattern in which more dorsal regions project more basally in both cochleas. The ipsilateral projections of lateral OC neurons appear to connect regions with similar best frequencies but the projections of medial OC neurons do not. Summation of data from all of the injections in each zone indicates that lateral OC projections are relatively evenly distributed throughout the ipsilateral cochlea but are predominantly to the apex in the contralateral cochlea. Medial OC projections are predominantly to the middle and basal parts of the cochlea on both sides with contralateral projections somewhat more basal than ipsilateral projections.

Differential olivocochlear projections from lateral versus medial zones of the superior olivary complex.

An anterograde tracer (35S-methionine) was injected unilaterally in the superior olivary complex (SOC) at regions previously demonstrated by retrograde labeling to contain olivocochlear (OC) cell bodies. Quantitative analysis of cochlear autoradiographs from these cats demonstrates that there are two OC systems. The lateral OC system has cell bodies lateral to the medial superior olivary nucleus (MSO) and projects to the inner hair cell (IHC) region bilaterally (mostly ipsilaterally). The medial OC system has cell bodies medial, ventral, and anterior to the MSO and projects to the outer hair cell (OHC) region bilaterally (mostly contralaterally). A single medial OC neuron innervates many small patches of OHCs with substantial gaps between the patches. Medial OC neurons also appear to project to the IHC region to a small extent. A review of the literature with the medial-lateral division of OC efferents in mind reveals many differences between these two systems. In particular, lateral OC axons are unmyelinated and innervate the dendrites of radial afferent fibers under IHCs, whereas medial OC axons are myelinated and directly innervate OHCs. Although both systems appear to be cholinergic, the lateral OC system also shows met-enkephalin-like immunoreactivity. The synapses of the medial OC system are formed in development before those of the lateral OC system and they degenerate more slowly after the OC axons are cut. The many differences between these two OC systems suggest that they are functionally separate systems.

Is there a medial nucleus of the trapezoid body in humans?

The medial nucleus of the trapezoid body (MNTB) appears to be a prominent auditory structure in many mammals. However, the presence of an MNTB in the human brain has not been clearly established. One of the most characteristic features of the cat MNTB is the presence of large somatic endings with multiple synaptic sites, the calyces of Held. We examined adult human brains at both light and electron microscopic levels and found neurons with unusually large endings in a location that is similar to that for the MNTB in other animals. Moreover, the sizes and shapes of some cells in this area are similar to the principal cells of the cat MNTB. These observations support the idea that humans have cells that resemble MNTB neurons in other species. It has been suggested that the cat MNTB may be involved in the generation of wave 3 of its brainstem auditory evoked potentials, so the presence of an MNTB in the human brain may have implications in interpreting brainstem potentials in man.

Tonotopic organization of the anteroventral cochlear nucleus of the cat.

A quantitatively accurate map of the tonotopic organization of the anteroventral cochlear nucleus (AVCN) was derived from single unit recordings. Histologically localized single unit recordings from many animals were mapped onto a computerized atlas of the cochlear nucleus, and surfaces of constant characteristic frequency (CF) estimated with the aid of computer graphics. In anterior AVCN the surfaces of constant CF were found to be parallel planes, whereas in posterior AVCN they progressively deviated from this simple description. A further complication was noted in the most posterior portion of the AVCN where units with very different CF was found in close proximity. Comparison of the tonotopic map with descriptions of cellular organization shows conclusively that different CF ranges are dominant in the various cytoarchitectonic regions of the AVCN.

A block model of the cat cochlear nucleus.

A three-dimensional block model of the cochlear nucleus of the cat was constructed from histologic sections. Boundaries of various subdivisions, based on cytoarchitectonic criteria, were included in the model. Usage of the block model in correlating physiological and anatomical data is illustrated by localizing characteristic waveforms of gross evoked responses and characteristic frequencies of single units.

Single unit activity in the posteroventral cochlear nucleus of the cat.

Single unit activity in the posteroventral cochlear nucleus (PVCN) was recorded for a variety of stimulus conditions. The units were classified according to their response characteristics. The locations of units were plotted onto a three-dimensional block model of the cochlear nucleus. Certain types of units that responded best to the onsets of stimuli were located predominantly in the octopus cell region of the PVCN. The remainder of the PVCN, which contains a rather heterogeneous collection of small and multipolar cells, was found to contain several types of units with the dominant type being "chopper" units.

Single unit activity in the dorsal cochlear nucleus of the cat.

Single unit activity was examined in three component layers of the dorsal cochlear nucleus (DCN): the molecular layer, the fusiform cell layer, and the polymorphic layer (deep DCN). Electrophysiological units were classified into types on the basis of their activity under a variety of stimulus conditions. In the molecular layer spike activity was small and difficult to isolate. Almost all units in the fusiform cell layer could be classified as either "pauser" or "buildup" units. Classification of units in the deep DCN was sometimes difficult, but "pauser," "chopper," and some "on" units were found. The "on" types of units tended to be located in the more superficial part of the deep DCN. Unit locations were referred to a three-dimensional block model of the cochlear nucleus.