Morphometry of intracellularly labeled neurons of the auditory nerve: correlations with functional properties

Axon termination conditions for electrical stimulation

The cable model for electrical stimulation near the terminal of a passive fiber is derived for excitation by an arbitrary, time-varying, applied extracellular field. Unless the termination impedance is comparable to that of mammalian node of Ranvier, the end-conditions require the longitudinal intracellular current at the fiber terminal to be negligibly small. This requirement substantially alters the membrane potential profile from that obtained with a fiber of infinite length. Stimulation near the end of a fiber may result in lower thresholds and may reverse the anodal/cathodal threshold ratio obtained with stimulation in the mid-portion of the fiber. Chronaxie for stimulation near the terminal may be much smaller than at a distance from the terminal and the strength-duration curve may be nonmonotonic. These differences may have significant implications for any application of electrical stimulation where fiber terminations may play a role in the excitatory process.

Postnatal maturation of human spiral ganglion cells: light and electron microscopic observations

The presence of two types of ganglion cells, based on cell size and other morphologic parameters, is well established in the adult mammalian and human spiral ganglion. On the other hand, there is little data concerning cell morphology in the neonatal spiral ganglion. The present study was undertaken to evaluate the differences in the morphometry and distribution of cell types in the spiral ganglion of the human neonate as compared to the adult. A total of five human temporal bones from two neonates and three infants were included in this study. Light microscopic analysis of all specimens was performed, and electron microscopic evaluation of a 14 day old neonatal spiral ganglion was accomplished. The segmental density of spiral ganglion cells was higher in the neonate than in the adult. The prevalence of type II spiral ganglion cells was higher in the neonate than has been reported in the adult, particularly in the middle and apical turns where type II cells constituted 24% and 26% of all ganglion cells, respectively. The prevalence of type II ganglion cells decreased with age, particularly in the middle and apical turns. In the neonate, the maximal cross sectional area of type I neurons increased from the base to the apex and seemed to increase with age especially in the basal turn. The present study strongly supports a clear differentiation of type I and type II ganglion cells in the human neonate and that the prevalence of type II cells is greater in the neonate than the adult. This finding is discussed with reference to postnatal development of the spiral ganglion.

Pathophysiology of the glutamatergic synapses in the cochlea

The synapses between the inner hair cells (IHCs) and the radial auditory dendrites are thought to be glutamatergic. Besides its fast excitatory properties, glutamate is known to be neurotoxic when released in excess or incompletely recycled. In the cochlea, this may occur in two pathological conditions: ischemia and noise trauma. We have further investigated the acute excitotoxicity (i.e. the swelling of type I afferent dendrites) by electron microscopy processing on guinea pig cochleas after an ischemic exposure lasting 5 to 40 min. The radial auditory dendrites reacted to ischemia in a time-dependent manner, with the swelling extending when the duration of ischemia increased. The type and the specificity of swelling were comparable to what acutely occurs after an exposure to glutamate analogs such as kainic acid or AMPA. A protection against this swelling was obtained by perfusing the cochlea with glutamate antagonists prior to ischemia. DNQX, an antagonist at AMPA/kainate receptors, had a powerful protective effect, and almost complete protection was obtained by perfusing both DNQX and D-AP5 (a NMDA antagonist). The latter results indicate that the two classes of glutamate receptors (AMPA/kainate and NMDA), both found to be electrophysiologically active at the IHC-auditory nerve synapse, are also involved in the excitotoxic processes. In addition, we also report data involving dopamine (its D2 agonist piribedil) a putative neurotransmitter at the lateral efferent synapses, in a postsynaptic protection of primary auditory neurons during transient ischemia. Altogether, these findings constitute a promising pharmacological approach of cochlear pathologies such as neural presbycusis.

Electrically evoked auditory brainstem response to stimulation of different sites in the cochlea

It is often assumed that stimulation of the eighth nerve with brief current pulses results in a highly synchronous discharge of action potentials from the ensemble of excited nerve fibers. Latency data derived from the electrically evoked auditory brainstem response (EABR) of the guinea pig and human indicate that there is a baso-apical gradient of response latencies in both species. Place-specific stimulation of different regions of the guinea pig cochlea was achieved by positioning a narrowly spaced bipolar electrode into each of the four spiral turns. Peak latencies of EABR waves I and III both revealed longer latencies with increasing distance from the base. Different amplitude growth functions were obtained from each turn. Place-specific stimulation in humans implanted with the Nucleus 22-electrode device was achieved by presenting current pulses across electrodes spaced 1.5 mm apart. A gradient of wave V latencies was observed in all subjects studied. The presence of a latency gradient would influence results obtained in efforts to correlate evoked potential amplitude measures with nerve survival. It may also provide information useful in characterizing the surviving fiber population in human implant subjects.

Central projections of auditory nerve fibers of differing spontaneous rate, II: Posteroventral and dorsal cochlear nuclei

Response properties of auditory nerve fibers (ANFs), including threshold sensitivity, vary systematically with spontaneous discharge rate (SR) (Liberman, M.C.: J. Acoust. Soc Amer. 63:442-455, 1978). Thus, an understanding of the mechanisms underlying signal transformation in the cochlear nucleus (CN) must include a description of any SR-based difference in ANF projections. This study is the second of a pair describing the CN projections of intracellularly labeled ANFs of known SR, the first of which summarized projection to the anteroventral CN (Liberman, M.C.: J. Comp. Neurol. 313:240-258, 1991). For each swelling from each labeled fiber, the position (within CN subdivisions), the size, and the type of cell contacted (if determinable) was noted: roughly one in four labeled swellings appeared in intimate contact with the soma or proximal dendrites of a CN cell. In all such cases, cell size and swelling size were measured. As reported for auteroventral cochlear nucleus, the ANF innervation of the small-cell regions of posteroventral CN (PVCN) was almost exclusively by low- and medium-SR fibers. Other significant SR-based trends in ANF projections included 1) a tendency for high-SR fibers to contact larger cells in PVCN, 2) a meager projection of low- and medium-SR fibers to octopus cells, and 3) a tendency in the dorsal CN (DCN) for low-SR terminals to end closer to the fusiform cell layer than high-SR terminals. There were no significant SR-based difference in ANF swelling sizes in any subdivision. A consideration of the average cell sizes, ANF swelling sizes and estimated numbers of ANFs of different CF and SR converging on each CN cell help explain some of the differences in response transformation associated with different cell types in the CN.

Performance of the avian inner ear

In spite of morphological similarities, the avian inner ear has apparently developed mechanisms of sound transduction that differ from the mammalian solution. This paper is a compilation of the present knowledge.

Innervation of the amphibian and basilar papillae in the leopard frog: reconstructions of single labeled fibers

Amphibians have two auditory organs specialized for reception of airborne sounds: the amphibian papilla and the basilar papilla. In this report we examine the morphology of the ganglion cells and the afferent innervation of the sensory epithelium in both auditory organs of the leopard frog, Rana pipiens pipiens. Extracellular injections of either biocytin or horseradish peroxidase (HRP) were made into the VIII nerve; they labeled ganglion cells, their axons, and their terminal fibers within the papillae. Ganglion cells that projected to either the amphibian papilla or basilar papilla had cell bodies that were morphologically distinct from other labeled cells. In the amphibian papilla thick fibers terminated in the rostral portion and thin fibers terminated in the caudal portion. Labeled fibers in the rostral portion traveled short distances before making contacts with up to nine hair cells whereas labeled fibers in the caudal portion traveled longer distances and contacted no more than five hair cells. In the basilar papilla labeled fibers were thick (around 4 microns) and terminated on as many as nine hair cells. Consistent with studies from the bullfrog, Rana catesbeiana, our results suggest that the amphibian papilla of R. pipiens pipiens has a convergent innervation (i.e., multiple hair cells provide input to a single ganglion cell) and is topographically organized. However, in contrast to reports in other ranid species, a highly convergent innervation like that found in the amphibian papilla is also found in the basilar papilla.

Topographic organization of the cochlear spiral ganglion demonstrated by restricted lesions of the anteroventral cochlear nucleus

The morphological organization of the central projections of the cat cochlear spiral ganglion into the cochlear nucleus has been investigated by creating restricted lesions in the anteroventral cochlear nucleus (AVCN) in order to ablate selectively either the lateral or the medial aspect of isofrequency projection laminae. Such lesions induced highly selective retrograde degeneration of spiral ganglion cells. Ablation of the lateral part of the AVCN resulted in degeneration of cells within the scala tympani portion of the ganglion, whereas medial lesions within the AVCN induced degeneration of the scala vestibuli portion of the ganglion. Since most, if not all, of the primary afferent axons of the cochlear nerve bifurcate into ascending and descending branches as they enter the brainstem, it is noteworthy that selective damage to the ascending branch in the AVCN was sufficient to induce retrograde degeneration of the spiral ganglion cell somata. The peripheral and central axons also degenerated, and the losses of both the radial nerve fibers in the osseous spiral lamina and the central axons passing into the modiolus displayed selective topographies that paralleled the cell loss within the spiral ganglion. The results of this study support our previous hypothesis, based upon earlier horseradish peroxidase labeling experiments, that there is a topographic organization to the projection of the spiral ganglion within the isofrequency laminae that is orthogonal to the frequency representation within the ventral cochlear nuclei (VCN). That is, in addition to the spiral frequency organization of the ganglion, represented by the dorsal-to-ventral frequency map in the VCN, there is also an orderly and sequential distribution of inputs from the vertical (scala tympani-to-scala vestibuli) dimension of the spiral ganglion across the lateral-to-medial axis of the VCN. The interaction of these two topographic representations, distributed across the three dimensions of the VCN, must partly define the selective and/or integrative neuronal response properties at this first level of central nervous system processing of auditory signals within the cochlear nuclei.

Spatial organization of the auditory nerve according to spontaneous discharge rate

Auditory-nerve fibers in mammals have been classified into three functional subclasses according to spontaneous discharge rate (SR). In cat, the peripheral terminals of these SR groups are segregated around the sensory cell circumference (Liberman, '82, Science 216:1239-1241). The present study shows that this spatial segregation is at least partly maintained through the peripheral axonal course from sensory cell to spiral ganglion. Analysis of intracellularly labeled auditory-nerve fibers shows that peripheral axons and cell bodies of low- and medium-SR fibers tend to be found closer to scala vestibuli than high-SR fibers. Since low- and medium-SR fibers tend to be thinner, this SR-based segregation can also be demonstrated as a fiber-caliber gradient in the osseous spiral lamina. The issue of SR-based spatial segregation is relevant to reports that ganglion cells near scala vestibuli project to different regions of the cochlear nucleus than cells near scala tympani (Leake and Snyder, '89, J. Comp. Neurol. 281:612-629). Combining the results of the two studies suggests that there may be some SR-based spatial segregation of inputs to the cochlear nucleus.

Round-window recorded potential of single-fibre discharge (unit response) in normal and noise-damaged cochleas

Unit responses (URs) of eighth-nerve fibres have been determined at the round window by spike-triggered averaging in both normal and pathological guinea pig cochleas. The pathology was mainly noise-induced damage. The URs have been analysed with respect to their dependence on the fibre's threshold, characteristic frequency (CF) and spontaneous rate (SR). The results from normal cochleas confirmed earlier data (Prijs, 1986): the UR has a diphasic waveform and the amplitude of its negative first peak is about 0.1 microV. From the six parameters (amplitude, latency, and width of the two peaks) by which the UR was described only the amplitude of the positive peak showed a significant variation with CF: a small decrease with increasing CF (CF-range 0.1 to 20 kHz). This finding may possibly be caused by oscillations in the spike-triggered average for low CFs. URs for most low- and medium-SR fibres were found to be large (greater than 0.3 microV). However, this result is interpreted as an artefact caused by synchrony of fibre spontaneous activity. In damaged cochleas only slight changes of the UR were found: the waveform duration became significantly shorter and on some occasions the positive peak increased in amplitude, but latency and amplitude of the negative component of the UR remained unchanged.

Phase-locking of auditory-nerve discharges to sinusoidal electric stimulation of the cochlea

The activity of auditory-nerve fibers was recorded in anesthetized cats in response to sinusoidal electric stimuli applied through a bipolar electrode pair inserted about 5 mm into the cochlea through the round window. The synchronization index was calculated from period histograms for frequencies ranging from 0.2 to over 10 kHz. The stimulus artifact was largely eliminated through the use of differential micropipettes and an adaptive digital filter. Measured synchronization indices were many times larger than the indices that could be attributed to the residual stimulus artifact. Synchronization indices at each stimulus frequency varied considerably from fiber to fiber, even in the same animal. The dependence of synchrony on stimulus frequency was also variable, decreasing monotonically in some fibers and nonmonotonically in others. The average electric synchronization index for all fibers did not fall as steeply with frequency as does the average synchrony for acoustic stimuli. The finding of significant phase locking to electric stimuli well above 1 kHz suggests that the poor frequency discrimination of cochlear-implant recipients for single-channel stimulation above this frequency may be due to the inability of the central processor to make effective use of the available phase-locking information for monaural stimulation.

Central projections of auditory-nerve fibers of differing spontaneous rate. I. Anteroventral cochlear nucleus

Auditory nerve fibers have been subdivided into three functional groups (Liberman, M.C. [1978] J. Acoust. Soc. Am. 63:442-455) differing in acoustic sensitivity and spontaneous discharge rate (SR). Using intracellular injection of horseradish peroxidase, the present study analyzes the projections of these three neuronal subclasses to the various subdivisions of the anteroventral cochlear nucleus (AVCN) and to the different cell types found therein. The average number of swellings and number of cells contacted decreased from low- to medium- to high-SR groups. However, these differences in terminal elaboration were not evenly distributed throughout the AVCN. The small cell cap was almost exclusively innervated by low- and medium-SR fibers, i.e., those with the highest acoustic thresholds. Within anterior AVCN, spherical-cell innervation was seen from all SR groups, whereas almost all multipolar cell innervation was from low- and medium-SR fibers. In the posterior AVCN, multipolar-cell innervation was equally likely from all SR groups, whereas globular cells were preferentially contacted by high-SR fibers. These SR-based trends in cochlear nucleus innervation help explain some of the known physiological properties of cell-types in each subdivision. They also suggest that additional physiological study of the small cell cap may be key in elucidating the functional significance of the low-SR population.

Postnatal maturation of spiral ganglion neurons: a horseradish peroxidase study

Using an in vitro cochlear preparation from postnatal hamsters, spiral ganglion cells (SGCs) were labeled retrogradely following extracellular injections of HRP into the cochlear nerve. In 24 cochleae from hamsters between postnatal days (P) 0 and 10, the neuronal morphology of 201 SGCs and their peripheral axons were analyzed. From P 0 to 3, labeled SGCs had few distinguishable features. Although SGCs could be traced separately to inner hair cells (IHCs) and outer hair cells (OHCs), they all had roughly bipolar-shaped cell bodies. Approximately half of the labeled SGCs had peripheral axons that spiraled some distance before entering radial fiber bundles. From P 3 to 7, SGCs increased in size by nearly 30% and the number of SGCs with spiraling peripheral axons decreased to near zero. At P 10, the central axon diameter to peripheral axon diameter ratios distinguished two populations of SGCs. The hair-cell innervation patterns of SGCs also changed morphologically as a function of postnatal age. At P 0, radial fiber (RF) terminals of peripheral axons contacted as many as 8 IHCs; by P 3, RFs contacted typically one or two IHCs. The terminal portions of peripheral axons contacting OHCs did not show any appreciable spiral until P 2. By P 5, individual outer spiral fibers (OSFs) had greater spiral lengths underneath row-3 OHCs and the number of OHC contacts was also greatest for row-3 OSFs. These data suggest that SGCs undergo a systematic maturational process. Furthermore, the morphological differentiation of SGCs occurs after they have established separate inner and outer hair cell innervations.

Analytical theory for extracellular electrical stimulation of nerve with focal electrodes. II. Passive myelinated axon

The cable model of a passive, myelinated fiber is derived using the theory of electromagnetic propagation in periodic structures. The cable may be excited by an intracellular source or by an arbitrary, time-varying, applied extracellular field. When the cable is stimulated by a distant source, its properties are qualitatively similar to an unmyelinated fiber. Under these conditions relative threshold is proportional to the cube of the source distance and inversely proportional to the square of the fiber diameter. Electrical parameters of the model are chosen where possible, from mammalian peripheral nerve and anatomic parameters from cat auditory nerve. Several anatomic representations of the paranodal region are analyzed for their effects on the length and time constants of the fibers. Sensitivity of the model to parameter changes is studied. The linear model reliably predicts the effects of fiber size and electrode-fiber separation on threshold of cat dorsal column fibers to extracellular electrical stimulation.

Unmyelinated axons of the auditory nerve in cats

This paper describes some central terminations of type II spiral ganglion neurons as labeled by extracellular injections of horseradish peroxidase (HRP) into the auditory nerve of cats. After histological processing with diaminobenzidine, both thick (2-4 microns) and thin (0.5 microns) fibers of the auditory nerve were stained. Whenever traced, thick fibers always originated from type I spiral ganglion neurons and thin fibers always from type II ganglion neurons. Because the labeling of type II axons faded as fibers projected into the cochlear nucleus, this report is limited to regions of the ventral cochlear nucleus near the auditory nerve root. The central axons of type II neurons are unmyelinated, have simple yet variable branching patterns in the cochlear nucleus, and form both en passant and terminal swellings. Under the light microscope, most swellings are located in the neuropil but they are also found in the vicinity of cell bodies, nodes of Ranvier of type I axons, and blood vessels. Eighteen en passant swellings in the neuropil were located by light microscopy and resectioned for electron microscopy; two of these swellings exhibited ultrastructural features characteristic of chemical synapses. The data indicate that inputs from outer hair cells might be able to influence auditory processing in the cochlear nucleus through type II primary neurons.

Synaptic connections of the auditory nerve in cats: relationship between endbulbs of held and spherical bushy cells

This report focuses on a class of large synaptic endings, the endbulbs of Held. These endings are located in the anteroventral cochlear nucleus and arise from the axons of type I spiral ganglion neurons. Axons were stained with horseradish peroxidase (HRP) using intracellular injections of single fibers or extracellular injections into the auditory nerve. Individual endbulbs or pairs of endbulbs that converged onto the same spherical bushy cell were examined with the aid of a light microscope and subjected to morphometric analyses. Endbulbs of fibers having low spontaneous discharge rates (SR, less than or equal to 18 spikes/sec) have a more complex shape than those of high SR fibers (greater than 18 s/s), a feature represented by systematic differences in endbulb silhouette perimeter without differences in endbulb silhouette area. Consequently, the ratio, silhouette area divided by silhouette perimeter, yields a "form factor" separating endbulbs of high SR from those of low SR. High SR fibers had ratios greater than 0.52 (mean = 0.63 +/- 0.09), whereas low SR fibers had ratios less than 0.52 (mean = 0.45 +/- 0.06). Pairs of endbulbs with unknown physiological properties had similar form factor values, despite the wide range of values observed in the endbulb population. These data imply that endbulbs converging upon the cell body of a spherical bushy cell arise from fibers of the same SR group. Electron microscopic examination was conducted on the endbulb of one physiologically characterized and intracellularly stained auditory nerve fiber (CF = 1.4 kHz; SR = 55 s/s) and its unstained endbulb mate with the aid of serial ultrathin sections. In addition to the well-known axosomatic synapses, these endbulbs formed axodendritic synapses: 11.7% for the HRP-labeled endbulb and 13.3% for the unlabeled endbulb. The axodendritic synapses appear to occur on dendrites of nearby spherical bushy cells and may represent a mechanism whereby single endbulbs can disperse activity to multiple neurons in the cochlear nucleus. We propose that axosomatic synapses preserve fiber SR groupings, whereas axodendritic synapses may not.

Degeneration of cochlear neurons as seen in the spiral ganglion of man

A light and electron microscopic study of the spiral ganglion cells of a normal individual, a patient with Meniere's syndrome, and a patient with long-standing unilateral profound deafness was done to evaluate patterns of neural degeneration in the human inner ear. Parametric data for the normal spiral ganglion are presented and compared with the pathologic ganglia. In the ear with Meniere's syndrome, the nuclear area and axonic diameter of spiral ganglion cells were significantly smaller than in the contralateral and the normal ear. This was interpreted as evidence of neuronal degeneration in Meniere's syndrome. In the spiral ganglion of the long-standing deaf ear, there appeared to be selective preservation of large cells with no dendritic processes, contrary to the pattern of degeneration seen in the spiral ganglion of the animal.

Origin of infrasound sensitive neurones in the papilla basilaris of the pigeon: an HRP study

Infrasound sensitive afferent fibres recorded in the pigeon cochlear ganglion were marked by intracellular injections of horseradish peroxidase (HRP). All stained fibres were found to innervate hair cells in the basilar membrane between 90 and 950 microns from its apical end. No fibres to the macula lagenae were found. Nine of the 10 completely stained fibres contacted hair cells located abneurally on the free basilar membrane, the tenth ended over the neural limbus near its abneural border. All fibres innervated between two and nine hair cells. This is in contrast to common auditory fibres in the bird that were reported to innervate only one hair cell located neurally over the neural limbus. This paper, therefore, demonstrates, for the first time, physiologically defined fibres that do not end on 'inner' hair cells.

On the pathophysiology of tinnitus; a review and a peripheral model

In this paper I investigate the consequences of the assumption that tinnitus is the result of correlated neural activity in auditory nerve fibers under 'no sound' conditions. Two possible pathological conditions capable of causing this correlation are ephaptic excitation of one nerve fiber by neighboring nerve fibers and synchronization of the various synapses in individual hair cells. The first condition is likely to be found in cases suffering from acoustic neuroma where the myelin sheath of the auditory neurons is damaged. The second condition is attributed to a spontaneous excess influx of K+ or Ca2(+)-ions into the hair cell resulting in transient hair cell depolarizations causing synchronous transmitter release at all hair cell synapses. This condition is postulated in noise trauma and ototoxic drug damage of the inner hair cell membrane. The model produces the excess of short interspike intervals found in auditory nerve fiber recordings in animal models of tinnitus as well as the theoretically required correlation in the activity of neighboring neurons.

Single-fibre responses to clicks in relationship to the compound action potential in the guinea pig

Poststimulus time histograms (PSTHs) to clicks of standard level were measured in eighth-nerve fibres of normal-hearing guinea pigs. In the context of studying the fibres' contribution to the compound action potential (CAP), the PSTHs are described with the parameters latency (tp), amplitude (Ap) and synchronization (Sp) of the dominant PSTH peak. These parameters are considered in relation to characteristic frequency (CF) and spontaneous rate (SR). An adequate description for tp is one in which tp is constant for non-phase-locking fibres (CF above 3 kHz) and it is an exponential function of CF for phase-following fibres. The low-SR fibres (SR below 5 spikes/s) had smaller amplitudes and longer latencies than the other ones. The variations of Ap with CF can be explained by the varying synchronization of the response.

Central projections of cochlear nerve fibers in the alligator lizard

The auditory (cochlear) ganglion cells of the alligator lizard (Gerrhonotus multicarinatus) give rise to two types of peripheral fibers: tectorial fibers, which contact hair cells covered by a tectorial membrane, and free-standing fibers, which contact hair cells without a tectorial membrane. To determine the central projections of these fibers, we applied intracellular and extracellular injections of horseradish peroxidase (HRP) to the peripheral component of the cochlear nerve. After histological processing with diaminobenzidine, individual cochlear nerve fibers could be traced through serial sections with the aid of a light microscope and drawing tube. The projection patterns formed two morphologically distinct groups. Neurons whose peripheral processes contacted tectorial hair cells in the cochlea projected to three divisions of the cochlear nucleus: nucleus magnocellularis lateralis (NML), nucleus magnocellularis medialis (NMM), and nucleus angularis lateralis (NAL). Neurons whose peripheral processes contacted free-standing hair cells projected primarily to the nucleus angularis medialis (NAM), although some also sent a single, thin branch to the NML; these neurons never projected to NAL or NMM. Morphometric comparisons of tectorial and free-standing fibers demonstrate that tectorial fibers have a larger axonal diameter, form a greater number of terminal swellings, and make proportionally more somatic contacts. By correlating the morphologically defined groups with previously reported physiologically defined groups, we conclude that different divisions of the cochlear nucleus are associated with separate frequency ranges and that stimuli in the different frequency ranges may be processed separately in the brain.

Morphometric analysis of normal human spiral ganglion cells

A morphometric analysis of the human spiral ganglion was performed at an ultrastructural level. Cells of the basal, middle, and upper middle turns were evaluated by serial section. The following parameters were evaluated: myelinization, area, diameter, circumference, and degree of roundness of the cell body and nucleus; diameter of the axon and dendrite in their initial segments; and process configuration. Analysis of variance suggested there were at least two types of cell bodies within the human spiral ganglion, best delineated by the dimensions of the cell body and nucleus and the ratio of diameter of the initial segments of axonic and dendritic processes. Myelinization and degree of roundness were relatively unimportant differential parameters. Cluster (multivariate) analysis of the six most important differential parameters suggested the possibility of five subgroups among the groups of large and small cells. For the middle and upper middle turns, cluster analysis suggested that there may be as many as three groups of cells based on morphometric analysis. These results are discussed in relation to other morphologic and physiologic data in the mammalian spiral ganglion.

Morphology and physiology of cells in slice preparations of the posteroventral cochlear nucleus of mice

In an effort to understand what integrative tasks are performed in the cochlear nuclei, the present study was undertaken to describe neuronal circuits in the posteroventral cochlear nucleus (PVCN) anatomically and physiologically. The cochlear nuclear complex receives auditory information from the cochlea through the auditory nerve. Within the cochlear nuclei, signals travel along several parallel and interconnected pathways. From the cochlear nuclei, transformed versions of the signals are passed to higher auditory centers in the brainstem. We have recorded electrophysiological responses from cells that were subsequently visualized with horseradish peroxidase (HRP). Responses to shocks to the auditory nerve root and to intracellularly injected current pulses were recorded and correlated with morphology. Two types of stellate cells and octopus cells were distinguished. T stellate cells project out of the cochlear nuclei through the Trapezoid body; D stellate cells do not. The axons of D stellate cells extend Dorsalward to the dorsal cochlear nucleus (DCN) but have not been traced out of the nucleus. Both T and D stellate cells have terminal collaterals in the multipolar cell region of the PVCN and in the DCN. The endings of one T stellate cell formed a narrow band rostrocaudally in the fusiform cell layer of the DCN that resembled an isofrequency band. The endings of one D stellate cell lay closely apposed to multipolar cells in the deep layer of the DCN. The dendrites of T stellate cells are often aligned along the path of auditory nerve fibers and end in tufts, whereas those of D stellate cells extend radially in the plane of the lateral surface of the PVCN toward granule cell areas and branch sparingly. Octopus cells have dendrites oriented perpendicularly to the path of auditory nerve fibers. Their axons were cut medially in the slices; none had collateral branches. Both T and D stellate cells were monosynaptically excited to threshold by shocks to the nerve root, indicating that they could participate in local circuits that we measure physiologically. T stellate cells have action potentials that peak at about 0 mV and are followed by single undershoots. The D stellate cell that was best impaled fired overshooting action potentials that were followed by double undershoots. Octopus cells were monosynaptically excited to threshold by shocks to the auditory nerve.

Ultrastructural studies of the peripheral extensions (dendrites) of type I ganglion cells in the cat

An ultrastructural study of peripheral extensions (dendrites) of type I ganglion cells in seven healthy adult cats and one cat that underwent bilateral cochlear injection of neomycin was undertaken. Morphologic evidence revealed that the peripheral process (dendrite) consistently has a smaller diameter than the central process. As the dendrite reaches the cell body, there is a constricted segment with a length that ranges from 10 to 30 microns, and a diameter of 0.5 microns. This region is covered by a continuous myelin sheath that does not thin. The central process (axon) does not have a corresponding constriction, and consists of a myelin covered, smoothly tapered segment that extends to the first axonal node of Ranvier. In the deafened cat, some cell bodies of the surviving ganglion cells appeared to have a residual portion of dendrite. These morphologic characteristics might have physiologic implications both in the mechanisms of normal hearing and in cochlear implantation.

Latencies of eighth nerve fibre responses with respect to their relative contribution to the compound action potential in the guinea pig

We present a study of the latencies of click-evoked post-stimulus time histograms (PSTHs) in the guinea pig in the context of the auditory nerve fibre's contribution to the compound action potential (CAP) recorded at the round window. The latencies of the dominant PSTH peak were studied as a function of relevant physiological fibre properties, in particular the characteristic frequency (CF) and the spontaneous discharge rate (SR). We found that high-SR fibres have shorter latencies than low-SR fibres. These findings are discussed in the context of correlation between synaptic morphology and SR as described in the literature. The PSTH latency as a function of CF is described separately for low- and high-CF fibres for each of the two SR subgroups. Finally, we discuss to what extent the various subgroups of fibres contribute to the N1 peak of the CAP, the most commonly studied component.

Spontaneous rates, thresholds and tuning of auditory-nerve fibers in the gerbil: comparisons to cat data

Characteristics of 245 auditory nerve fibers in eleven Mongolian gerbils are described in terms of spontaneous rates, thresholds, and tuning curves. The animals were reared in a low-noise environment and had similar hearing thresholds across frequency. Tuning curves were obtained with an algorithm developed to characterize the tuning of auditory fibers in cat, thereby allowing direct comparisons to published data from cat. The results demonstrate that basic similarities exist between gerbil and cat data, although some minor differences are also apparent. Tuning curve bandwidths, as measured 10 and 40 dB above the thresholds at the characteristic frequency (CF), follow trends found in cat data. Like cat, auditory nerve fibers in the gerbil have a range of spontaneous rates. In individual gerbils, fibers associated with low spontaneous rates have higher thresholds than do fibers of similar CF with high rates. Five of the eleven gerbils showed profiles of spontaneous rate across frequency reminiscent of those obtained from quiet-raised young cats. The profiles of the remaining gerbils tended to be compressed to a smaller range of spontaneous rates for characteristic frequencies above about five kHz, much like older cats with unknown noise histories. The cause of the spontaneous compression is not obvious. The correspondence between cat and gerbil with regard to spontaneous rate and CF threshold implies the presence of fundamental mechanisms that are common to mammalian auditory systems.

Morphology and response properties of single olivocochlear fibers in the guinea pig

Cochlear efferent and afferent units were recorded from the spiral ganglion in anesthetized guinea pigs. Efferent units were identified by their regular discharge patterns and their long response latencies to tone bursts. In some cases these physiological criteria were confirmed by anatomical tracing of single axons labeled with horseradish peroxidase. Labeled efferent axons traveled in the vestibular nerve root and the intraganglionic spiral bundle, and crossed the tunnel to innervate up to 61 outer hair cells over longitudinal distances of 0.15 to 2.1 mm. Efferent units were subdivided into classes by their excitatory responses to monaural tone bursts. Fifty-seven percent responded only to ipsilateral-ear sound, 28% only to contralateral-ear sound, and 15% to sound in either ear. Tuning curves from efferent units were generally similar in shape to afferent units, often having equally low thresholds and equally high Q10s. Efferent Q10s were somewhat lower from preparations anesthetized with a Urethane/Innovar combination than from preparations anesthetized with a Nembutal/Innovar combination. Efferent units with spontaneous activity were uncommon at the start of the recording sessions but were more frequently encountered later in the experiments. Spontaneous activity could often be suppressed by tonal stimuli, even sometimes to the non-excitatory ear of Ipsi and Contra units.

Temporal synchronization in the primary auditory response in the pigeon

Spike potentials were recorded from single fibres in the auditory nerve of the pigeon. In responses elicited by tonal stimuli, the timing of each spike relative to stimulus waveform was measured and period histograms were constructed. Phase locking of spikes was estimated in terms of a synchronicity index obtained by vector addition within the period histogram. A second measure of synchrony in the spike responses was obtained, that of temporal dispersion. For a population of fibres, vector strength of phase locking decreased for frequencies above 1 kHz, as reported for several other species. Temporal dispersion, however, also decreased with frequency, indicating enhanced temporal synchrony as frequency increased within the bandwidth of phase locking. The upper frequency limit of phase locking appears to depend on irreducible jitter of biological origin in the timing of spikes. For individual fibres, the bandwidth of synchronization of spikes consistently exceeds the response area, covering in addition the areas of suppression adjacent to the response area. Spike trains suppressed by a tonal stimulus become synchronized to that stimulus. Phase angles of synchronized responses systematically change as a function of tone level, when tone frequency is above or below CF, as reported for other avian species. Synchronicity and phase angle intensity functions are quite independent of spike rate intensity functions.

A model proposing synaptic and extra-synaptic influences on the responses of cochlear nerve fibres

A unique property of sensory coding in the vertebrate auditory system is the existence of the classical form of excitatory centre-inhibitory surround in relative spike rate along the stimulus frequency dimension, in addition to a representation of temporal fine structure of high frequency periodic stimuli in the discharge pattern of primary afferent spike trains. We present a model which designates three factors that influence rate and temporal synchrony in spike responses; an excitatory factor, a suppressive factor and a synchronizing factor. The model proposes that an essential integration of bioelectric signals occurs in the primary afferent fibre. It is presumed that mean spike rate depends on mean level of membrane depolarization and synchronization depends on periodic modulation of membrane potential at the spike initiating zone. In the model, the excitatory factor is synaptically-mediated, excitatory post-synaptic potential (e.p.s.p.); the suppressive factor is negative DC polarization of the fibre membrane and the synchronizing factor is AC modulation of the fibre membrane potential. It is proposed that both the negatively-polarizing and high-frequency modulating signals are derived from extracellular current flow in the cochlea.

Central trajectories of type II spiral ganglion neurons

Previous attempts to trace the central pathways of the thin axons from type II spiral ganglion neurons have been hampered by technical difficulties such as fading of the reaction product as distance increases from the injection site (Ryugo et al.: Soc. Neurosci. Abstr. 12:779, '86; Brown: J. Comp. Neurol. 260:591-604, '87). By using small rodents (gerbils and mice), which have short auditory nerves, we have succeeded in filling the entire central axon and terminals of type II neurons after peripheral injections of horseradish peroxidase. The general course of the type II fibers within the auditory nerve and cochlear nucleus is similar to that of type I fibers except that terminals from type II neurons are often found in regions of the cochlear nucleus that have high densities of granule cells.

Brainstem branches from olivocochlear axons in cats and rodents

Horseradish peroxidase was used to label axons of olivocochlear (OC) neurons by intracellular injections in cats and extracellular injections in rodents. These axons arise from cell bodies in the superior olivary complex and project to the cochlea. En route to the cochlea, the thick axons (greater than 0.7 micron diam.) of medial olivocochlear (MOC) neurons formed collaterals that terminated in the ventral cochlear nucleus, the interstitial nucleus of the vestibular nerve (in cats), and the inferior vestibular nucleus (in rodents). The thin axons (less than 0.7 micron diam.), presumed to arise from lateral olivocochlear (LOC) neurons, did not branch near the CN. Within the CN, the MOC collaterals tended to ramify in and near regions with high densities of granule cells, regions also associated with the terminals of type II afferent axons (Brown et al.: J. Comp. Neurol. 278:581-590, '88). These results suggest that those fibers associated peripherally with outer hair cells (MOC efferents and type II afferents) are associated centrally with regions containing granule cells, whereas those fibers associated with inner hair cells peripherally (LOC efferents and type I afferents) are not.

Physiological and psychophysical correlates of temporal processes in hearing

Discharges of auditory nerve fibers are synchronized to stimulus frequencies below 4-5 kHz. The phase-locking phenomenon has been studied in considerable detail in several animal species. Although strikingly close correspondences exist between phase-locking behavior in animals and human perceptual performance on certain tasks, there is still no clear evidence that the human brain actually bases perceptual decisions on temporally encoded frequency information. The alternative to temporal coding is rate-place coding, in which frequency is assigned on the basis of peaks in cochlear excitation patterns. This paper reviews pertinent physiological, psychophysical and modeling data in three classes of experiment whose results are explanable in terms of both rate-place and temporal processing of neural responses. The experiments deal with the pitch of complex tones, vowel identification, and pure-tone frequency discrimination. The data described here suggest that temporal models of frequency coding compete well with and in some cases offer a more parsimonious explanation of perceptual performance than rate-place codes do, particularly at low and middle frequencies. A potentially important implication of the analyses conducted here is that humans may not code frequency information in synchronized activity as well as other species. The data suggest that within limits the human ear is capable of using either temporal and rate-place frequency codes, and that the specific code employed by the perceptual processor is task-dependent.

Central projections of intracellularly labeled auditory nerve fibers in cats: morphometric correlations with physiological properties

The central arborizations and endings of type I spiral ganglion neurons were labeled with intracellular injections of horseradish peroxidase (HRP) after their characteristic frequency (CF) and spontaneous discharge rate (SR) were physiologically determined. A fiber-by-fiber analysis was conducted and the morphological data compared with the fiber's response properties. The total number of branch points was correlated with total fiber length, a relationship that remained relatively constant when analyzing the ascending and descending branches together or separately. On the other hand, the ascending branches of four out of five fibers having CFs below 0.5 kHz bifurcated and gave rise to a pair of terminal endbulbs of Held. Low- and medium-SR fibers gave rise to more endings than did high-SR fibers, especially on the ascending branch. This difference was accounted for by small endings, a category composed of terminal boutons, string endings, and small complex endings. The categories of modified endbulbs, and endbulbs of Held did not vary in number with respect to fiber SR. The mean area of each ending type within the small ending category was statistically smaller for low- and medium-SR fibers than for high-SR fibers, whereas the mean area of modified endbulbs and endbulbs of Held was not correlated with fiber SR. Total ending area per fiber appeared independent of either CF or SR. These results are discussed in relation to issues of conservation of axon arborizations and terminals, and convergence of input from the different SR groups.

Afferent innervation of outer hair cells in adult cats: I. Light microscopic analysis of fibers labeled with horseradish peroxidase

Outer spiral fibers (OSFs), the afferent innervation of the outer hair cells (OHCs), were retrogradely labeled following horseradish peroxidase injections into the cat's auditory nerve. The peripheral branching patterns of 85 OSFs from adult cochleas were reconstructed. Fibers contacted OHCs via terminal or en passant swellings; however, the latter were seen exclusively in the apical half of the cochlea. Many OSFs also gave off branches ending on structures other than OHCs. Fibers in the cochlear apex were much more highly branched than in the base. Most fibers contacted only one row of OHCs, and more fibers contacted row 1 than row 2 or row 3 OHCs. Third-row fibers were the most highly branched in all cochlear regions. These results are consistent with a growing body of morphological evidence that suggests that the peripheral branching patterns of OSFs may be fundamentally similar in all mammalian ears.

Afferent innervation of outer hair cells in adult cats: II. Electron microscopic analysis of fibers labeled with horseradish peroxidase

Outer spiral fibers (OSFs) were retrogradely labeled with injections of horseradish peroxidase. After the peripheral arborization patterns were reconstructed at the light microscopic level (Simmons and Liberman, '88), restricted regions of selected fibers were analyzed via electron-microscopic reconstruction of serial sections. The ultrastructural data in the present study suggested that the contact between the outer hair cell (OHC) and the terminal swellings of OSFs corresponds to the afferent synapse described in numerous other ultrastructural studies. The en passant swellings that contacted OHCs also appeared to be points of synaptic contact. However, en passant synapses were not always associated with a swelling of the OSF at the point of contact: thus, the light-microscopic reconstructions probably underestimate the numbers of synapses. OSF branches terminating well below the OHCs were seen to end most commonly in intimate contact with the Deiters' cells. Membrane specialization was occasionally seen at this point of contact; however, the specialization was sufficiently undifferentiated to preclude identification.

A model of electrical excitation of the mammalian auditory-nerve neuron

A model of the mammalian auditory-nerve neuron has been developed based on the classical work of Frankenhauser and Huxley [(1964) J. Physiol. 171, 302-315], modified by McNeal [(1976) IEEE Trans. Biomed. Eng. BME-23, 329-336], and Reilly et al. [(1985) IEEE Trans. Biomed. Eng. BME-32, 1001-1011], and fine tuned to represent physiological data obtained from single auditory-nerve fiber experiments in squirrel monkeys. The model is capable of reproducing neural action potential waveforms due to electrical stimulation, and can reliably predict action potential thresholds and strength-duration curves. This paper explains the derivation of the mathematical model and the effects of varying certain independent parameters including fiber diameter, length of the nodes of Ranvier, internodal length, and myelin thickness. The model parameters were selected according to the anatomical findings of Liberman and Oliver [(1984) J. Comp. Neurol. 223, 163-176], and Liberman (Pers. Commun.). The length of the unmyelinated termination of the auditory-nerve that survives after aminoglycoside damage to the inner ear has not been experimentally determined. Therefore, it was investigated as an independent variable in the model. An unmyelinated terminal length of 10.0 micron was found to most accurately describe the experimental neural strength-duration curves obtained from aminoglycoside-deafened squirrel monkeys. The parameter that had the next most significant effect on the model was fiber diameter which affects all conduction pathways, across the membrane and through the fiber. Finally the results of the model are compared with behavioral data obtained from patients and monkeys implanted with cochlear prostheses. In the companion paper [(1987) Hear. Res. 31, 267-286] predictions of the model are quantitatively compared with single-neuron data from squirrel monkeys.

Auditory-nerve single-neuron thresholds to electrical stimulation from scala tympani electrodes

Single auditory-nerve neuron thresholds were studied in sensory-deafened squirrel monkeys to determine the effects of electrical stimulus shape and frequency on single-neuron thresholds. Frequency was separated into its components, pulse width and pulse rate, which were analyzed separately. Square and sinusoidal pulse shapes were compared. There were no or questionably significant threshold differences in charge per phase between sinusoidal and square pulses of the same pulse width. There was a small (less than 0.5 dB) but significant threshold advantage for 200 microseconds/phase pulses delivered at low pulse rates (156 pps) compared to higher pulse rates (625 pps and 2500 pps). Pulse width was demonstrated to be the prime determinant of single-neuron threshold, resulting in strength-duration curves similar to other mammalian myelinated neurons, but with longer chronaxies. The most efficient electrical stimulus pulse width to use for cochlear implant stimulation was determined to be 100 microseconds/phase. This pulse width delivers the lowest charge/phase at threshold. The single-neuron strength-duration curves were compared to strength-duration curves of a computer model based on the specific anatomy of auditory-nerve neurons. The membrane capacitance and resulting chronaxie of the model can be varied by altering the length of the unmyelinated termination of the neuron, representing the unmyelinated portion of the neuron between the habenula perforata and the hair cell. This unmyelinated segment of the auditory-nerve neuron may be subject to aminoglycoside damage. Simulating a 10 micron unmyelinated termination for this model neuron produces a strength-duration curve that closely fits the single-neuron data obtained from aminoglycoside deafened animals. Both the model and the single-neuron strength-duration curves differ significantly from behavioral threshold data obtained from monkeys and humans with cochlear implants. This discrepancy can best be explained by the involvement of higher level neurologic processes in the behavioral responses. These findings suggest that the basic principles of neural membrane function must be considered in developing or analyzing electrical stimulation strategies for cochlear prostheses if the appropriate stimulation of frequency specific populations of auditory-nerve neurons is the objective.

Morphology of labeled afferent fibers in the guinea pig cochlea

Cochlear afferent and efferent fibers in the guinea pig were labeled by focal extracellular injections of horseradish peroxidase into the spiral ganglion of the basal turn. The morphology and pattern of termination of these fibers were studied by light microscopy. Fibers labeled by injections into the peripheral side of the ganglion could be grouped on the basis of their courses and terminations in the cochlea into two classes of afferent fibers, two classes of efferent (olivocochlear) fibers, and other presumably autonomic fibers. This paper describes the characteristics of labeled afferent fibers and their parent ganglion cells. Peripheral afferent fibers were grouped into two major classes: thick (mean diameter 1.7 micron) radial fibers projecting in a primarily radial fashion from the spiral ganglion and terminating on single inner hair cells and thin (mean diameter 0.5 micron) outer spiral fibers that spiral basalward in the organ of Corti to terminate on outer hair cells, usually in one row. For outer spiral fibers, the number of outer hair cells contacted and the length of the terminal region depend on the row of outer hair cells contacted, with third-row fibers forming, on the average, the most extensive region of termination. Within the spiral ganglion, two types of ganglion cells could be distinguished: type-I ganglion cells of large size (mean soma area = 216 microns 2) with a ratio of central process diameter to peripheral process diameter greater than one and type-II ganglion cells of smaller size (mean soma area = 100 microns 2) and a central to peripheral process ratio near one. In three cochleae in which injections were made central to the ganglion, 11 type-I ganglion cells have been traced to radial fibers contacting inner hair cells and eight type-II ganglion cells have been traced to outer spiral fibers contacting outer hair cells. Thus the afferent innervation of the guinea pig cochlea is similar to the pattern described in other mammals, in which there is separate innervation of the inner and outer hair cells by the two types of ganglion cells. The central axons of both types of ganglion cells were traced individually through serial sections of a block of tissue containing the cochlea, the auditory nerve, and the cochlear nucleus. They followed similar courses in the auditory nerve, and the axons followed into the cochlear nucleus bifurcated in similar regions of the interstitial portion.(ABSTRACT TRUNCATED AT 400 WORDS)

Morphology of labeled efferent fibers in the guinea pig cochlea

Efferent axons to the guinea pig cochlea were labeled by extracellular injections of horseradish peroxidase into the intraganglionic spiral bundle within the spiral ganglion. The terminal fibers formed by these axons were classified according to their patterns of termination within the basal turn of the cochlea. A class of terminal fibers designated "autonomic" forms a highly branched plexus in the osseous spiral lamina but does not enter the organ of Corti. The termination of single autonomics includes blood vessels as well as areas of the osseous spiral lamina not adjacent to blood vessels. Two major classes of efferent axons from the olivocochlear bundle enter the cochlea by way of the vestibulocochlear anastomosis and terminate either in areas near inner hair cells (IHC efferents) or onto outer hair cells (OHC efferents). The IHC efferents have thin axons throughout their course within the cochlea and can be divided into two subclasses. The most numerous subclass of IHC efferents (unidirectional) enters the inner spiral bundle and turns to spiral in only one direction for less than 1 mm and then forms a discrete termination including many en passant and terminal swellings that are within both the inner and tunnel spiral bundles. A less common subclass of IHC efferents (bidirectional) bifurcates upon entry into the inner spiral bundle to send branches both apically and basally. These terminal fibers take spiral courses that are greater than 1 mm in extent, often course in the tunnel spiral bundle for a large portion of the spiral, and form terminals throughout their extended spiral course. None of the IHC efferent fibers send branches to cross the tunnel to innervate the outer hair cells. A second major class of olivocochlear fibers, OHC efferent fibers, forms large boutons on the outer hair cells, and although they sometimes spiral beneath the IHCs for some length, they do not give off terminals to this region. The OHC efferent axons are thick and myelinated as they enter the cochlea, and they branch near the spiral ganglion to form several terminal fibers. Some of these terminal fibers are thin as they travel from the intraganglionic spiral bundle across the osseous spiral lamina to the organ of Corti, whereas others are thick and obviously myelinated as far peripheral as the habenula.(ABSTRACT TRUNCATED AT 400 WORDS)

Hair cell innervation by spiral ganglion neurons in the mouse

Horseradish peroxidase (HRP) was injected extracellularly into the auditory nerve of adult mice so that the enzyme could infuse individual spiral ganglion neurons. Forty-two well-stained neurons were reconstructed through serial sections from their cell bodies to peripheral terminations in the organ of Corti with the aid of a light microscope and drawing tube. No neuron was observed to innervate both inner and outer hair cells (IHCs and OHCs). Previous observations from neonatal mammals that reported that IHCs and OHCs were innervated by the same neuron are thus presumed to describe a transient developmental phenomenon. Two populations of spiral ganglion neurons were determined on the basis of the differences in receptor innervation. The type I neurons innervated exclusively IHCs by way of thick (1-2 microns) radial fibers, whereas the type II neurons innervated only OHCs by way of thin (approximately 0.5 micron) outer spiral fibers. Certain features of the peripheral process in the vicinity of the cell body were highly correlated with fiber type. This pattern of separate innervation of IHCs and OHCs by type I and type II neurons, respectively, may represent the general plan of afferent organization for the adult mammalian cochlea.

Single fiber mapping of spatial excitation patterns in the electrically stimulated auditory nerve

Spatial maps of electrical excitation were constructed by comparing electrical threshold with acoustic CF for large populations of auditory nerve fibers in cats. Thresholds among fibers with the same CF varied by factors of 4 or more. Monopolar electrodes, both intracochlear and extracochlear, excited fibers throughout the cochlea without spatial selectivity. Stimulation with intracochlear bipolar electrodes produced a minimum in the threshold distribution adjacent to the electrodes. With longitudinally oriented pairs, the width, depth, and location of the minimum shifted with stimulus polarity; spread of excitation throughout the cochlea occurred with stimulus intensities 6.2 to 14 dB above the lowest threshold. With radially oriented pairs, minima were sharper and deeper; spread of excitation occurred at intensities 23.7 to 32.8 dB above the minimum threshold.

Chronic ultrastructural changes in acoustic trauma: serial-section reconstruction of stereocilia and cuticular plates

Two cochleas with permanent, noise-induced threshold shifts of 40 to 60 dB (as measured by responses of single, auditory-nerve fibers) were analyzed in detail, first at the light-microscopic level, and subsequently with transmission electron microscopy of serial sections. The light-microscopic analysis showed that there was little hair cell loss, but widespread damage to the stereocilia, especially those on the inner hair cells and first-row outer hair cells. Transmission electron microscopy revealed no pathology in any cells or organelles in the organ of Corti except for the stereocilia and their rootlets. Stereocilia tufts on first-row OHCs and IHCs were badly damaged; those on second- and third-row OHCs appeared ultrastructurally normal. Within the IHC tuft, the damage to the tall, outer row of stereocilia was often selective: the shorter rows could remain ultrastructurally normal even when the tall row was completely missing. The data suggested that most of the structures which appear normal in a careful light-microscopic analysis are also normal at the ultrastructural level. This strengthens earlier suggestions that the correlations between light-microscopic stereocilia changes and alterations in single-unit physiology are causal in nature. The most common stereocilia pathologies were fracture, attenuation or complete loss of the stereociliary rootlets, especially in the region of the cuticular plate near the endolymphatic surface of the cell. The degree and extent of these changes were well correlated with the degree and extent of stereocilia disarray. Abnormalities of the actin-filament matrix within the stereocilia were extremely rare in unfused stereocilia, however, they were common when the stereocilia were part of a fusion bundle. Fusion of stereocilia was always associated with ectopic supracuticular cytoplasm. Based on the ultrastructural observations, different sequences of structural changes preceding the generation of disarray, loss or fusion of stereocilia are suggested.

Hair cell damage produced by acoustic trauma in the chick cochlea

Examination of pure-tone acoustic damage in the chick basilar papilla revealed that the location and extent of hair cell damage was a function of both the stimulus intensity and the age at which the chicks were exposed. Scanning electron microscopic evaluation of noise-exposed cochleae at post-hatching days 1, 10 and 30 permitted the identification of discrete regions of damage, including hair cells with stereocilia injuries as well as those lost from the epithelium. The hair cell damage was tonotopically distributed along the cochlea according to frequency. However, for each exposure frequency two distinct sites of damage were often produced, and their locations were correlated with stimulus intensity. At low intensities, a longitudinal strip of hair cell damage ran along the superior edge of the basilar papilla. As exposure intensity increased, a second damage site developed along the inferior edge of the basilar papilla, distal to the longitudinal strip. This second type of damage initially took the form of a series of laterally-oriented wedges, but at higher intensities, the wedges coalesced to form a large crescent-shaped patch of damage. The location of the damage sites for each frequency did not shift with age. However, there were differences in the extent and position of the damage which could be correlated with stimulus intensity and with changes in middle ear admittance during development [(1983) Development of Auditory and Vestibular Systems, pp. 3-25. Editor: R. Romand. Academic Press, New York]. These results suggest that developmental changes in the location and extent of hair cell damage depend on the effective stimulus intensity reaching the cochlea, rather than on alterations in the frequency coding of the hair cells.

The central projections of intracellularly labeled auditory nerve fibers in cats: an analysis of terminal morphology

The axons of physiologically characterized spiral ganglion neurons (type I) were stained throughout their arborizations in the cochlear nucleus by the intracellular injection of horseradish peroxidase (HRP). The tips of the axonal branches were marked by distinct swellings, ranging in size and shape from small boutons to large perisomatic ramifications. Electron microscopic analysis of such swellings revealed ultrastructural features characteristic of primary auditory synapses, consistent with the hypothesis that terminal swellings identifiable in the light microscope represent presynaptic endings. On the basis of light microscopic differences in size, these endings were organized into three categories. Endings of relatively small size (terminal boutons, free endings, boutons with filopodia, string endings, and small complex endings) composed 94% of all terminal endings. Within this category of small endings, there were predictable variations in relative size and regional distribution that related to the spontaneous discharge rate (SR) of the fiber. The endings of low and medium SR fibers (SR less than or equal to 18 spikes/second) were smaller on average than those of high SR fibers (SR greater than 18 spikes/second). Furthermore, there were more endings arising from the ascending branch than from the descending branch when comparing fibers of the low and medium SR group with those of the high SR group. There were not, however, obvious morphological features of this ending category that correlated with the characteristic frequency (CF, the pure tone frequency to which the neuron is most sensitive). A second category contained medium-sized complex endings, most of which formed axosomatic contacts. This category composed 4% of the population and was found in close proximity to the perikarya of globular, octopus, and spherical cells. The endings from low and medium SR fibers were smaller on average than those from high SR fibers. These endings did not vary in their parent branch distribution with respect to fiber SR, nor did they exhibit morphological features that correlated with fiber CF. The third category contained large complex endings (endbulbs of Held) and composed 2% of the ending population. Within the anteroventral cochlear nucleus, these large, complex endings made axosomatic contact with spherical cells in the anterior division and with globular cells in the posterior division. There were no systematic variations in ending size or branch distribution that correlated with fiber SR. There was, however, a relationship between ending size and fiber CF such that fibers having CFs below 4 kHz gave rise to the largest endbulbs.

Representation of a low-frequency tone in the discharge rate of populations of auditory nerve fibers

Discharge rates for populations of single auditory nerve fibers in response to 1.5 kHz tone bursts were measured in anesthetized cats. Separate plots of average rate vs. best frequency (rate-place profiles) were made for high, medium and low spontaneous rate (SR) auditory nerve fibers. At the lowest sound levels studied (34 dB SPL), all three SR groups show a peak in the rate-place profile centered around 1.5 kHz. At the highest sound levels studied (87 dB SPL), the average rates of the high and medium SR fibers are saturated across a wide range of best frequencies, but a peak around 1.5 kHz is maintained in the rate-place representation of the low SR fibers. These results show that in addition to the temporal information present in the discharge patterns of auditory nerve fibers, a rate-place representation of a single low-frequency tone exists in the auditory nerve over a wide range of sound levels.

Physiology and anatomy of single olivocochlear neurons in the cat

A surgical approach to the cat's VIIIth nerve has been developed which allows recordings to be made from efferent fibers of the olivocochlear bundle (OCB) as well as primary afferent fibers, without compromising the acoustic responsiveness of either ear. The designation of OCB fibers as those with regular interspike intervals was confirmed in five cases by intracellular labeling with horseradish peroxidase. Labeled fibers could be traced centrally to somata in the superior olivary complex and peripherally to large endings on outer hair cells. The locations of the labeled neurons are consistent with a classification as cells of the medial olivocochlear system [Warr and Guinan (1979): Brain Res. 173, 152-155]. Within the cochlea, efferent neurons branched profusely to innervate as many as 84 outer hair cells over as much as 2.8 mm of the organ of Corti. Efferent units had tuning curves which were similar to those of primary afferents, although most were somewhat more broadly-tuned than afferents from the same animal. The cochlear region innervated by an efferent neuron was always close to the place where afferent fibers of the same characteristic frequency (CF) would be found. Most efferents (89%) were excited by only one ear and showed no spontaneous activity. Neurons with an ipsilateral response (n = 3) had cell bodies in the contralateral brainstem and vice versa (n = 2). Binaural units (none of which were labeled) often had spontaneous discharge and were generally restricted to low-CF regions. Differences between low- and high-CF units, which cut across the monaural/binaural distinction, were seen in the dynamic range and minimum latency. Interanimal differences seen in the responses of the efferent neurons may be related to differences in the level of anesthesia.

Rapid adaptation depends on the characteristic frequency of auditory nerve fibers

When a tone burst is presented to the ear, the firing rate of single auditory nerve fibers is initially high, then rapidly declines toward a smaller value. The rate of this decline in gerbil is a complex function of stimulus frequency and intensity, and fiber characteristic frequency.