Summary. The functional role of GABA and glycine in monaural and binaural signal analysis was studied in single unit recordings from the central nucleus of the inferior colliculus (IC) of horseshoe bats (Rhinolophus rouxi) employing microiontophoresis of the putative neurotransmitters and their antagonists bicuculline and strychnine.Most neurons were inhibited by GABA (98%; N= 107) and glycine (92%; N = 118). Both neurotransmitters appear involved in several functional contexts, but to different degrees.Bicuculline-induced increases of discharge activity (99% of cells; N= 191) were accompanied by changes of temporal response patterns in 35 % of neurons distributed throughout the IC. Strychnine enhanced activity in only 53% of neurons (N= 147); cells exhibiting response pattern changes were rare (9%) and confined to greater recording depths. In individual cells, the effects of both antagonists could markedly differ, suggesting a differential supply by GABAergic and glycinergic networks.Bicuculline changed the shape of the excitatory tuning curve by antagonizing lateral inhibition at neighboring frequencies and/or inhibition at high stimulation levels. Such effects were rarely observed with strychnine.Binaural response properties of single units were influenced either by antagonization of inhibition mediated by ipsilateral stimulation (bicuculline) or by changing the strength of the main excitatory input (bicuculline and strychnine).

1. Previous studies of the superior olive of echolocating bats suggest that the lateral superior olive (LSO) retains the same structure and function as in other mammals but that the medial superior olive (MSO) is different in structure and possibly also in function. The present study is an examination of this idea in Pteronotus parnellii, a bat that has a large and well-defined MSO. 2. Using pure tones presented via earphones, we obtained data on frequency tuning for 60 single units and 96 multiunits in LSO and 94 single units and 154 multiunits in MSO. Of these we also obtained binaural response characteristics from 55 single units in LSO and 72 single units in MSO. 3. LSO and MSO each have a complete tonotopic representation, arranged in a sequence similar to that of other mammals studied. However, in both LSO and MSO there is an expanded representation of the frequencies around 60 kHz, the main frequency component of the bat's echolocation call; there is another expanded representation of the range around 90 kHz, the third harmonic of the call. The expansion of these frequency ranges suggests that the functions of LSO and MSO in Pteronotus are related to echolocation behavior. 4. The binaural characteristics of cells in LSO were essentially the same as those seen in other mammals. Most LSO units (93%) were excited by the ipsilateral ear and inhibited by the contralateral ear. The responses of nearly all LSO units were completely suppressed when the sound level at the two ears was equal. 5. The binaural characteristics of cells in MSO were different from those in nonecholocating mammals. Most MSO units (72%) were excited by the contralateral ear but were neither excited nor inhibited by the ipsilateral ear. Of the remaining units, 21% were excited by the contralateral ear and inhibited by the ipsilateral ear, and only 6% were excited by both ears. 6. The temporal discharge patterns of units in MSO differed from the tonic response pattern seen in LSO. Most MSO units had phasic response patterns, with a few spikes at the onset or offset of the stimulus; the response often changed from ON to OFF depending on stimulus frequency. 7. The results support the idea that in evolution LSO has remained unchanged, whereas MSO has undergone adaptation. The function of LSO in Pteronotus seems to be identical to that in other mammals, i.e., analysis of interaural sound level differences to derive azimuthal location. The function of MSO in Pteronotus must be different from that in nonecholocating mammals.(ABSTRACT TRUNCATED AT 400 WORDS)

A comparative study of the immunostain to antibodies directed against glutamic acid decarboxylase (GAD) and gamma-aminobutyric acid (GABA) in the ascending auditory pathway was carried out in horseshoe bats (Rhinolophus rouxi) and mustached bats (Pteronotus parnellii). In both species GAD/GABA-positive puncta (presumed axonal boutons) and GAD/GABA-positive cells were found in the cochlear nucleus, the superior olivary complex, the nuclei of the lateral lemniscus the inferior colliculus, and the medial geniculate body. General features of the immunostaining pattern in the auditory pathway agree with observations in other mammals. Quantitative analysis of puncta distribution shows that many auditory centers are characterized by subregional differences in puncta density and distribution. This indicates local differences in putatively inhibitory input related to connectivity and tonotopic organization. The following species characteristic features were found: 1) The dorsal non-laminated portion of the dorsal cochlear nucleus in horseshoe bats lacks the GAD/GABA-immunoreactive cells typical for the ventral laminated portion and the dorsal cochlear nucleus of other species. Clearly, a cytoarchitectonic specialization is accompanied by a loss of putatively GABAergic local inhibitory circuits. 2) The ventral division of the medial geniculate body of the mustached bat lacks GAD/GABA-immunopositive cells. Such cells are present in the horseshoe bat and other mammals. This finding implies functional differences in the organization of the medial geniculate body within the same mammalian order.

No abstract

Neurons of the columnar region of the ventral nucleus of the lateral lemniscus of Eptesicus fuscus respond with high-precision constant-latency responses to sound onsets and possess remarkably broad tuning. To study the synaptic basis for this specialized monaural auditory processing and to elucidate the excitatory or inhibitory nature of the input and output circuitry, we have used classical transmission electron microscopy, and postembedding immunocytochemistry for gamma aminobutyric acid (GABA) and glycine on serial semithin sections. The dominant putatively excitatory perisomatic input is provided by large calyx-like terminals that possess round synaptic vesicles and asymmetric synaptic contacts. Additionally, calyces contact the dendrites of neighboring neurons. Putatively inhibitory small boutons possess pleomorphic or flattened synaptic vesicles and symmetrical contacts and are sparsely distributed on somata and dendrites. Almost all neurons are glycine-immunoreactive. There is a moderate amount of glycine-immunoreactive puncta; GABA-immunoreactive puncta are rare. This suggests that (1) there is a fast robust excitatory synaptic input via calyx-like perisomatic endings, (2) calyx-like endings distribute frequency-specific excitatory input across isofrequency sheets by virtue of parallel synapses to somata and adjacent dendrites, and thus, dendritic integration may contribute to the broadening of frequency tuning, (3) the columnar region forms an inhibitory glycinergic feedforward relay in the ascending auditory pathway, a relay that is probably involved in creating filters for time-varying signals.

Adult mustached bats employ Doppler-sensitive sonar to hunt fluttering prey insects in acoustically cluttered habitats. The echolocation call consists of 4-5 harmonics, each composed of a long constant frequency (CF) component flanked by brief frequency modulations (FM). The 2nd harmonic CF component (CF2) at 61 kHz is the most intense, and analyzed by an exceptionally sharply tuned auditory system. The maturation of echolocation calls and the development of Doppler-shift compensation was studied in Cuba where large maternity colonies are found in hot caves. In the 1st postnatal week, infant bats did not echolocate spontaneously but could be induced to vocalize CF-FM signals by passive body motion. The CF2 frequency emitted by the smallest specimens was at 48 kHz (i.e., 0.4 octaves lower than the adult signal). CF-FM signals were spontaneously produced in the 2nd postnatal week at a CF2 frequency of 52 kHz. The CF2 frequencies of induced and spontaneous calls shifted upward to reach a value of 60.5 kHz in the 5th postnatal week. Standard deviations of CF2 frequency were large (up to +/-1.5 kHz) in the youngest bats and dropped to values of +/-250 Hz at the end of the 3rd postnatal week. Some individuals in the 4th and 5th postnatal weeks emitted with adultlike frequency precision of about +/-100 Hz. In the youngest bats, the 1st harmonic CF component (CF1) was up to 22 dB stronger than CF2. Adultlike relative levels of CF1 (-28 dB relative to CF2) were reached in the 5th postnatal week. In spontaneously emitted CF-FM calls, the duration of the CF2 component gradually increased with age from 5 ms to maximum values of 18 ms. Durations of the CF2 component in induced calls averaged 7 +/- 2.6 ms in the 1st postnatal week and 8.2 +/- 1.5 ms in the 5th postnatal week. There were no age-related changes in duration of the terminal FM sweep (3 +/- 0.4 ms) in both induced and spontaneous calls. The magnitude of the terminal FM sweep in spontaneous calls was not correlated with age (mean 13.5 +/- 2 kHz). Values for induced calls slightly increased with age from 11 +/- 2 to 13 +/- 2 kHz. The emission rate of induced CF-FM signals increased with age from values of 2.5 +/- 2 to 17 +/- 5 pulses/s. Values for spontaneously emitted calls were 4.4 +/- 3 and 9 +/- 4.5 pulses/s, respectively. Doppler-shift compensation, as tested in the pendulum task, emerged during the 4th postnatal week in young bats that were capable of very brief active flights, but before the time of active foraging outside the cave.

The ascending projections of the cochlear nucleus (CN) and the sources of descending inputs to the CN were investigated in horseshoe bats (Rhinolophus rouxi) by tracing the anterograde and retrograde transport of horseradish peroxidase (HRP or WGA-HRP) injected into the CN. The tracer was iontophoretically deposited into physiologically characterized regions of the cochlear nucleus (Feng and Vater, '85). We report the course and termination of pathways arising from the anteroventral (AVCN), posteroventral (PVCN), and dorsal (DCN) cochlear nucleus. The projection fields within the auditory brainstem centers (superior olivary complex [SOC]; lateral lemniscus complex [LLC]; and inferior colliculus [IC]) and their tonotopic organization according to the frequency representations at the injection sites are described. While the projection pattern is generally in accordance with other mammals, several species-characteristic features are noted: i) the lateral superior olive (LSO) receives tonotopically organized input from both the AVCN and PVCN; ii) the CN-projections to medial nuclear groups of the SOC located between the LSO and the medial nucleus of the trapezoid body do not support previously suggested homologies; iii) the ventral nucleus of the LLC can be subdivided into two divisions with distinct input patterns from the AVCN and PVCN, respectively.

The frequency-place map of the cochlea of mustache bats was constructed by the analysis of HRP-transport patterns in spiral ganglion cells following iontophoretic tracer injections into cochlear nucleus regions responsive to different frequencies. The cochlea consists of 5 half turns (total length 14.3 mm) and the representation of certain frequency bands can be assigned to specific cochlear regions: The broad high frequency range between 70 and 111 kHz is represented in the most basal half turn within only 3.2 mm. This region is terminated apically by a distinct narrowing of the scala vestibuli that coincides with a pronounced increase in basilar membrane (BM) thickness. The narrow intermediate frequency range between 54 and 70 kHz is expanded onto 50% of cochlear length between 4.0 and 11.1 mm distance from apex. The frequency range around 60 kHz, where the tuning characteristics of the auditory system are exceptionally sharp, is located in the center of this expanded BM-region in the second half turn within a maximum of innervation density. These data can account for the vast overrepresentation of neurons sharply tuned to about 60 kHz at central stations of the auditory pathway. In the cochlear region just basal to the innervation maximum, where label from injections at 66 and 70 kHz was found, a number of morphological specializations coincide: the BM is maximally thickened, innervation density is low, the spiral ligament is locally enlarged, and the 'thick lining', a dense covering of the scala tympani throughout the basal halfturn, suddenly disappears. Low frequencies up to 54 kHz are represented within the apical half turns over a 4 mm span of the basilar membrane. The data are compared to the cochlea of horseshoe bats and the possible functional role of the morphological discontinuities for sharp tuning and the generation of otoacoustic emissions is discussed.