Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Signal generation in the lateral superior olive : the rate code of interaural disparities of sound Adam, Trudy Jean 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-564908.pdf [ 11.81MB ]
Metadata
JSON: 831-1.0089656.json
JSON-LD: 831-1.0089656-ld.json
RDF/XML (Pretty): 831-1.0089656-rdf.xml
RDF/JSON: 831-1.0089656-rdf.json
Turtle: 831-1.0089656-turtle.txt
N-Triples: 831-1.0089656-rdf-ntriples.txt
Original Record: 831-1.0089656-source.json
Full Text
831-1.0089656-fulltext.txt
Citation
831-1.0089656.ris

Full Text

Signal Generation in the Lateral Superior Olive: The Rate Code of Interaural Disparities of Sound by Trudy Jean Adam B.Sc. , The University o f Calgary, 1990 M . A . , The University of British Columbia, 1992 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E S T U D I E S Neuroscience We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A January 2000 © Trudy Jean Adam, 2000 U B C Special Collections - Thesis Authorisation Form Page 1 of 1 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t .the U n i v e r s i t y o f B r i t i s h .Columbia, I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by the head i o f my department o r by h i s o r her r e p r e s e n t a t i v e s , i t i s u n d e r s t o o d t h a t c o p y i n g ' o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n ot 'be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f A / ^ g / y y j The U n i v e r s i t y o f B r i t i s h . C o l u m b i a Vancouver, Canada Date S e F 5 C O http://www.library.ubc.ca/spcoll/thesauth.html 9/18/00 T. J. A D A M i i 1. A B S T R A C T The azimuthal location of a sound source imposes interaural intensity disparities (IIDs) on binaural sound, which are encoded by principal neurons of the lateral superior olive (LSO). These disparities are thought to be represented in the chopper discharge pattern, characterized by regular repetitive firing at a precisely timed onset. Presumably, the integration o f ipsilateral excitation and contralateral inhibition determines the chopper rate code o f IID. It was hypothesized that this rate code must remain stable during prolonged stimulation to provide an effective localization code. Extracellular recordings of L S O chopper units during dichotic tonal stimulation revealed not only that short-term adaptation occurs in both the ipsilateral and contralateral inputs, but that binaural responses are stable only when the IID is zero. Therefore, the chopper rate code o f IID cannot provide a sufficient or reliable index o f azimuthal location. The L S O chopper pattern was hypothesized to arise partly from the interplay of intrinsic membrane properties o f the principal neuron. Whole-cell recordings o f L S O neurons in brainstem slices during direct current injection revealed that membrane properties involving early peak polarizations emphasized and accelerated response onset, and contributed to chopper response generation. A transient depolarizing potential was sensitive to 600 n M tetrodotoxin, 50 u M nickel, and 3 m M cesium, indicating contributions from persistent sodium, transient low-threshold calcium, and hyperpolarization-activated cation conductances. Potassium conductances sensitive to 4-aminopyridihe shaped the decay o f the transient potential. T w o inward (anomalous) rectifiers sensitive to 0.2 m M barium and 3 m M cesium, respectively, accounted for rectification in the hyperpolarized range. Interestingly, this repertoire o f conductances is similar to that proposed to support the preservation o f temporal information in phase-locking neurons. This implies that chopper responses may encode temporal aspects of the auditory stimulus. T. J. A D A M i i i The chopper pattern may not provide a simple rate code of IID, since it alone does not distinguish a moving sound source from a prolonged stationary one. Further, a specialized repertoire of intrinsic membrane conductances may allow the chopper neuron to encode intensity/IID dynamics o f the complex stimulus. Future research endeavours employing more complex stimulation paradigms wi l l certainly illuminate the exciting issue of sound coding in the L S O . T. J. A D A M iv TABLE OF CONTENTS 1. ABSTRACT II 2. LIST OF TABLES VIII 3. LIST OF FIGURES IX 4. INTRODUCTION 1 4.1 Overview: Audition in the natural environment 1 4.2 Physical cues for sound localization 2 4.3 Interaural disparity coding and sound localization in the auditory brainstem...5 4.3.1 Auditory brainstem circuitry 5 4.3.1.1 The medial superior olive (MSO) 6 4.3.1.2 The lateral superior olive (LSO) 7 4.3.2 Presynaptic specializations of olivary input neurons 7 4.3.3 Postsynaptic specializations of olivary input neurons 9 4.3.4 Models of binaural integration 10 4.3.4.1 The medial superior olive 11 4.3.4.2 The lateral superior olive 12 4.3.5 Temporal response properties of auditory brainstem neurons 14 4.3.5.1 V C N inputs to the superior olive 14 4.3.5.2 Principal neurons o f the L S O 15 4.3.6 Intrinsic membrane properties of auditory brainstem neurons 16 4.3.6.1 Phase-locking neurons of the auditory brainstem 17 4.3.6.2 Chopper neurons of the auditory brainstem 18 4.4 Implications of specializations for the LSO rate code of IID 19 4.5 The lateral olivocochlear system 21 4.5.1 LOC cytoarchitecture and circuitry. 21 4.5.2 LOC physiology and function 23 5. RESEARCH OBJECTIVES 24 (^EXAMINATION OF SHORT-TERM ADAPTATION IN THE SUPERIOR OLTVE26 6.1 Methods 28 6.1.1 Selection of animals. 28 6.1.2 Surgical preparation for acute electrophysiological experiments 28 6.1.3 Stimulus generation and delivery 29 6.1.4 Extracellular recordings 30 6.1.5 Paradigms addressing short-term adaptation 31 6.1.6 Assessment of short-term adaptation 33 6.1.7 Histological verification of neuron location 34 T . J . A D A M v 6.2 Results , 35 6.2.1 Location of recordings within the superior olive 35 6.2.2 Spike discharge of IE units during ipsilateral and contralateral stimulation 36 6.2.2.1 Responses to ipsilateral excitation 36 6.2.2.2 Responses to binaural stimulation 37 6.2.3 Short-term adaptation of excitation and inhibition in binaural LSO units 38 6.2.3.1 Ipsilateral excitation 39 6.2.3.2 Contralateral inhibition 41 6.2.3.3 Binaural responses 44 6.2.4 Post-stimulatory effects of short-term adaptation in IE units 47 6.2.4.1 Equal IIDs 47 6.2.4.2 Unequal IIDs 49 6.2.5 The influence of short-term adaptation on response onset latency. 51 7. EXAMINATION OF INTRINSIC MEMBRANE PROPERTIES IN NEURONS OF THE LSO 56 7.1 Introduction 56 7.2 Methods 58 7.2.1 Selection of animals 58 7.2.2 Surgical preparation of rat auditory brainstem slices 59 7.2.3 The recording chamber and associated equipment. 61 7.2.4 Solutions and electrodes 62 7.2.4.1 External solution 62 7.2.4.2 Internal solution for sharp microelectrodes 62 7.2.4.3 Internal solution for patch electrodes 63 7.2.5 Intracellular recordings : 63 7.2.5.1 Sharp microelectrodes 64 7.2.5.2 Patch electrodes 65 7.2.6 Current clamp recordings 65 7.2.6.1 Basic membrane properties 66 7.2.6.2 Spike discharge patterns 66 7.2.6.3 Synaptic potentials 67 7.2.7 Ion channel antagonism 67 7.2.8 Spike discharge characterization 68 7.2.8.1 Histogram formulation 68 7.2.8.2 Basic firing statistics 69 7.2.8.3 Discharge regularity 70 7.2.9 Neurobiotin injection and related histology. 71 7.3 Results 72 7.3.1 Two electrophysiological types of LSO neuron are distinguished by morphology 72 7.3.2 Action potential discharge in LSO principal neurons 75 7.3.3 Chopper-like firing patterns in the tissue slice: 76 T. J. ADAM v i 7.3.4 Discharge statistics for chopper-like patterns in the tissue slice 78 7.3.4.1 Relation of chopper-like discharge to membrane properties 81 7.3.5 Voltage dependence of membrane rectification in LSO principal neurons 83 7.3.5.1 Chopper-like firing 84 7.3.5.2 Membrane rectification in the depolarized voltage range 86 7.3.5.3 Rectification in the hyperpolarized voltage range 87 7.3.5.4 The linearity of current-voltage relations 89 7.3.6 Conductances contributing to chopper membrane dynamics. 90 7.3.6.1 Conductances shaping the depolarizing voltage sag 90 7.3.6.2 Conductances contributing to the depolarizing afterpotential 93 7.3.6.3 Conductances supporting the transient potential 95 7.3.6.4 Conductances contributing to the hyperpolarizing voltage sag 98 7.3.6.5 Conductances contributing to high threshold spikes 100 7.3.6.6 Outward conductances modulating chopper spike discharge 101 7.3.7 Chopper-like discharge following current pre-pulses 102 7.3.8 Synaptic potentials. 104 8. THE DESCENDING EFFERENTS OF THE LATERAL OLIVOCOCHLEAR SYSTEM 106 8.1 Action potential discharge in LOC efferents of the LSO 106 8.1.1 Basic membrane properties and subthreshold responses 106 8.1.2 The temporal pattern and discharge statistics of LOC efferent firing 107 9. DISCUSSION .. 110 9.1 Chopper-like responses are produced by LSO principal neurons 110 9.2 Generation of the chopper pattern in the LSO 112 9.2.1 Regulation of the onset spike 112 9.2.2 Regulation of repetitive firing 114 9.2.3 Spike rate accommodation 116 9.2.4 Membrane rectification 118 9.3 Comparison with other auditory neurons in vitro 119 9.3.1 VCNstellate neurons. /19 9.3.2 VCN bushy neurons, and MNTB principal cells 121 9.4 Mechanisms underlying sensitivity to HDs 123 9.4.1 The integration of excitation and inhibition in the LSO 123 '9.4.2 The relative timing of excitation and inhibition during IID coding 125 9.5 Implications for the representation of HD in the LSO 127 9.6 The lateral olivocochlear efferents of the LSO . 130 9.6.1 Summary ofprimary findings 130 9.6.2 Identification of delay neurons as efferent lateral olivocochlear neurons 130 9.6.3 Functional significance. 7 32 10. CONCLUSIONS 133 T . J . A D A M vi i 11. N O M E N C L A T U R E 136 12. R E F E R E N C E S 139 T. J. A D A M vii i 2. LIST OF T A B L E S 5.01 Ion equilibrium potentials for whole cell patch recordings at 22 °C. 160 5.02 Basic electrophysiological properties distinguishing chopper and delay neurons in recordings with sharp microelectrodes conducted at 34 °C. 161 5.03 Basic electrophysiological properties distinguishing chopper and delay neurons in whole cell patch recordings conducted at 22 °C. 162 T. J. A D A M ix 3. LIST OF FIGURES 3.01 The auditory brainstem in transverse section 164 4.01 Horseradish peroxidase marks within the superior olivary complex 166 4.02 Effects o f contralateral inhibition on ipsilateral excitation 168 4.03 Short-term adaptation in ipsilateral excitation of a binaural L S O unit 170 4.04 Short-term adaptation in contralateral inhibition of a binaural unit 172 4.05 Short-term adaptation in binaural responses o f a typical L S O unit 174 4.06 Recovery from short-term adaptation for equal interaural intensities 176 4.07 Recovery from adaptation for unequal interaural intensities 178 4.08 Onset latencies for binaural responses during recovery 180 4.09 Short-term adaptation and recovery of excitation in a monaural L S O unit 182 5.01 The recording chamber employed for in vitro recordings 184 5.02 Morphology of chopper neurons and delay neurons i n the L S O 186 5.03 Temporal pattern of spike discharge for chopper neurons 188 5.04 Spike discharge statistics for chopper neurons 190 5.05 Intrinsic properties o f chopper neurons in the L S O 192 5.06 Chopper spike discharge shows voltage dependence 194 5.07 Voltage dependence of chopper neuron intrinsic membrane properties 196 5.08 Voltage dependence o f the chopper neuron depolarizing afterpotential 198 5.09 Anomalous rectification contributes to the depolarizing afterpotential 200 5.10 The depolarizing afterpotential is mediated in part by calcium 202 5.11 Anomalous rectification and a low threshold calcium conductance yie ld the D A P 204 5.12 A subthreshold sodium conductance contributes to the transient potential 206 5.13 A low threshold calcium conductance contributes to the transient T. J . A D A M x potential 208 5.14 The low threshold calcium conductance contributes to the onset spike 210 5.15 Anomalous rectification contributes to the nickel sensitive transient potential 212 5.16 A-type potassium conductances sculpt the decay of the transient potential 214 5.17 A-type potassium conductances contribute to steady-state depolarizations 216 5.18 Other sustained potassium conductances limit depolarizations 218 5.19 Chopper neurons exhibit slowly inactivating high threshold calcium spikes 220 5.20 A-type potassium conductances truncate chopper responses 222 5.21 Sustained potassium conductances contribute to chopper discharge transience 224 5.22 Intrinsic properties mediate chopper sensitivity to the membrane potential 226 5.23 Synaptic depression in inhibitory postsynaptic potentials 228 5.24 Stability o f excitatory postsynaptic potentials 230 6.01 Intrinsic properties of delay neurons in the L S O 232 6.02 Temporal patterns of spike discharge for the delay neurons 234 6.03 Spike discharge statistics for delay neurons_l 236 T. J. A D A M 1 4. I N T R O D U C T I O N 4.1 Overview: Audition in the natural environment In natural hearing, the listener is continuously confronted with sounds originating from many sources in his/her surroundings. The separation of these sounds into discrete meaningful representations in the brain is a difficult perceptual problem. Sounds from competing sources arrive at the ears intermixed and overlapping in frequency spectra. In order to achieve accurate and reliable auditory perception, the listener must first disentangle stimulus components, and then associate them with discrete physical events (Bregman, 1990). The separation of multiple sound signals is thought to be based on the unique locations of sound sources in space around the listener. Neural encoding of source location in the horizontal plane around the listener (azimuth) is thought to provide a substrate for sound segregation. O n the basis o f neural localization codes, listeners are presumably able to attend to sounds from a particular source in a noisy background. The physical cues employed in sound localization are well understood. A s well , a substantial research base identifies single unit response properties that could serve as localization codes for the separation of sound signals. However, the utility o f these purported codes during prolonged stimulation, as in natural listening, remains to be demonstrated. This is especially true for high-frequency sounds (over - 1 . 2 kHz) . The purpose of this dissertation research was to (i) determine the reliability of the localization code for high frequencies during prolonged stimulation, and (ii) delineate intrinsic mechanisms contributing to the generation of this code in the auditory brainstem. In the following introductory sections, neural mechanisms underlying azimuthal sound localization are reviewed, and their implications for auditory signal extraction discussed. T w o series o f experiments are then reported. The first explores the behaviour of the putative localization code in the lateral superior olive during prolonged stimulation. T. J. A D A M 2 The second delineates intrinsic membrane properties that may support the preservation of the localization code in the lateral superior olive. 4.2 Physical cues for sound localization Sound localization participates in auditory signal separation by providing a reliable code o f each sound source's unique location along the azimuth. O n the basis of common localization representations, the central auditory system is presumably able to dissect the auditory environment into coherent and meaningful perceptions, even though sounds arriving at the ears overlap in frequency spectra and time. Psychophysical research into sound localization supports this role for binaural signal processing, since removal o f binaural cues renders speech comprehension in noisy environments much more difficult (Gelfand, Ross, and Mi l le r , 1988). This is presumably due to the elimination of interaural disparities as a basis for stimulus separation. To follow is a review o f the physical cues thought to be employed in sound localization. Azimuthal sound localization is thought to be achieved through the detection and coding of interaural disparities in binaural sound. Due to the physical separation o f the two ears, sound waves arrive earlier, and are more intense, at one ear than the other. Resultant interaural differences in time and intensity o f sound are employed in the azimuthal localization of binaural sound (Searle, Braida, et al., 1976; Lewis , 1983; Irvine, 1986). v The delay in arrival o f sound to one ear, relative to the other, imposes interaural time differences (ITDs) on the binaural signal, which are maximal when the sound source is directly in front of one ear, and minimal at midline. Both the difference in time-of-arrival, and ongoing phase differences in the fine structure of periodic signals provide information for sound localization. Changes in I T D along the azimuth follow a monotonic function from zero to approximately 800 us in humans (Mi l l s , 1958), and approximately T. J. A D A M 3 350 us in cats (Roth, et al., 1980). Therefore, ITDs provide a direct representation of azimuthal location, increasing with more lateral locations. The distance between the ears also imposes differential attenuation o f the signal's intensity (interaural intensity disparity, IID). Diffraction o f the sound signal by the head, neck, and pinnae, results in an acoustic shadow over the far ear, and a resultant reduction of signal intensity reaching that ear. Generally, IIDs follow a monotonic function with azimuthal location, being maximal when the sound source is located directly in front of one ear, and minimal at midline. In humans, this range covers IIDs of 0 to approximately 25 dB SPL as sound moves from midline to directly in front of one ear (Shaw, 1974). However, IID is a complex function of azimuthal location and pinna amplification effects (Phillips, and Gates, 1982), particularly for animals used in localization studies. Therefore, the relation o f IID to free-field azimuthal location is complicated. The effectiveness of ITDs and IIDs in sound localization is not straightforward. Each cue can be ambiguous, depending on the spectral content of the signal. For high frequencies, interaural phase differences become ambiguous as they approach and exceed the period o f the signal. In humans, these ambiguities arise for wavelengths over 800 us, and for cats, they arise for wavelengths over 350 us. These values correspond to signal frequencies o f 1.2 and 2.8 k H z , respectively. In the case of IIDs, low frequency signals are reflected less than high ones. Consequently, the acoustic shadow cast by the head and shoulders is reduced for low frequencies. Therefore, IIDs are not as prominent, and are ineffectual as perceptual cues for sounds lower than approximately 1.4 k H z in humans (Kuhn, 1977; Irvine, 1986; Domnitz, and Colburn, 1977; Phillips, and Gates, 1982). The frequency dependent dichotomy in effectiveness of localization cues has led to the development of the duplex theory of sound localization. The central hypothesis is that high frequency sounds are localized on the basis o f IIDs, while low frequencies are localized according to ITDs (Jeffress, 1948; Stevens, and Newman, 1936; Y i n , and Chan, 1988). Early researchers reported that localization accuracy was frequency dependent, with errors being largest between 1.5 and 2.5 k H z in humans (Stevens, and Newman, T. J. A D A M 4 1936; M i l l s , 1958). This corresponds well with the frequency dependence of interaural disparities, since neither IIDs nor ITDs would be ideal indices of sound source location in the intermediate frequency range. While the frequency ranges o f effectiveness for these cues varies across species, with head size and prominence o f the pinnae in hearing, the duplex theory o f sound localization has generally held up to scientific scrutiny. Experimental support has been gained from studies employing dichotic stimulation o f the ears, where IIDs and ITDs can be manipulated selectively. The demonstration of time-intensity trading was critical in this regard. A time difference favouring one ear could be offset by an intensity disparity favouring the other ear. This finding alone has confirmed the fundamental tenets o f the duplex theory o f sound localization (Domnitz, and Colburn, 1977; Durlach, and Colburn, 1978; Joris, and Y i n , 1995; Joris, 1996; Joris, and Y i n , 1998). Employment of complex or broadband stimuli in dichotic listening studies may also be interpreted in the context o f the duplex theory. Humans are capable o f localizing complex high frequency sounds on the basis o f ITDs in the low frequency stimulus envelope (Henning, 1974; McFadden, and Pasanen, 1976; Nuetzel, and Hafter, 1976). Al so , broadband stimuli (e.g. white noise) are more accurately localized than pure tones. It has been assumed that this is due to the increased availability of interaural disparity cues in the complex stimulus for localization (Oldfield, and Parker, 1984; Terhune, 1974; Stevens, and Newman, 1936). Clearly, interaural disparities do provide a reliable means by which to localize sound, but their relative importance depends heavily on the spectral content of the stimulus. Psychophysical findings have clear implications for a role of ITDs and IIDs in sound localization and sound signal extraction from the noisy environment. The duplex theory of sound should be supported by a neural substrate for sound localization, which would be revealed by electrophysiological study o f auditory neurons during dichotic stimulation of the ears. Research addressing the anatomical and physiological bases for T. J . A D A M 5 sound localization has focused in the first site of binaural interaction, the auditory brainstem. 4.3 Interaural disparity coding and sound localization in the auditory brainstem The auditory brainstem features as the primary site of physiological investigations in sound localization due to the recognition that stimulus separation must be performed very rapidly in order to support the behavioural implications of hearing (e.g. predation). Therefore, this process must be accomplished at the early stations o f the central auditory system. The auditory brainstem is the first site of binaural information (IID and I T D ) processing in the ascending auditory system. We now review the role of the auditory brainstem in sound localization by first considering the circuitry involved, and then turning our attention to the physiological response properties supporting interaural disparity codes. 4.3.1 Auditory brainstem circuitry The auditory brainstem mediates information transfer between the periphery (auditory nerve) and the midbrain (primarily the inferior colliculus). Historically, it has been conceptualized as a network of several parallel pathways, each serving to encode a unique aspect of the stimulus. Even at higher centres, notably the inferior colliculus (IC), these circuits appear to remain functionally separated (Roth, A i tk in , et al., 1978; Calford, and Ai tk in , 1983; Warr, 1982). The most prominent parallel circuits o f the brainstem involve the medial and lateral superior olives ( M S O and L S O ) , and comprise the earliest sites of binaural integration in the ascending auditory system. Each of these nuclei receive auditory input from each ear, and integrate information to produce a code o f interaural disparities present in the binaural signal. T. J. A D A M 6 The input pathways for the superior olives are relatively simple and well delineated. Bo th begin as the eighth cranial nerve (auditory root) enters the cochlear nucleus, and fibers bifurcate to reach the anterior and posterior divisions of the ventral nucleus ( A V C N and P V C N ) . One branch terminates in the A V C N . The other innervates the P V C N en route to the dorsal cochlear nucleus (Fekete, Rouiller, et al., 1984; Feldman, and Harrison, 1969; Osen, and Roth, 1969). These projections are similar for the M S O and L S O brainstem circuits. 4.3.1.1 The medial superior olive (MSO) The M S O circuit originates in the A V C N , where large spherical bushy cells project to the ipsilateral M S O , terminating on primary projection neurons, the principal cells (Warr, 1966; Cant, and Casseday, 1986; Cant, and Morest, 1979b; Schwartz, 1984; Brawer, Morest, and Kane, 1974; Smith, 1995). From the contralateral A V C N , large spherical bushy cell axons also provide direct innervation to the M S O , crossing midline in the trapezoid body. In all areas o f each A V C N , the projection pattern is tonotopic, and the converging inputs to the M S O are tonotopically matched (Osen, and Roth, 1969; Osen, 1970; Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972; Goldberg, 1969). Both the ipsilateral and contralateral projections terminate on the M S O principal cells via excitatory glutamatergic synapses (Warr, 1972; Cant, and Casseday, 1986; Smith, 1995). Primary fibers entering the M S O to innervate the principal neurons do so as collaterals of projections to other areas of the SOC and higher auditory centres (Morest, 1968; Warr, 1966). Axons of principal M S O neurons ascend through the lateral lemniscus to terminate in the ipsilateral central nucleus of the inferior colliculus (ICc), with a small output to the contralateral IC (Ai tk in , and Schuck, 1985; Adams, 1979; Brunso-Brechtold, Thompson, and Masterton, 1981). T. J. A D A M 7 4.3.1.2 The lateral superior olive (LSO) The S-shaped L S O is also innervated by both the ipsilateral and contralateral A V C N s , but input from the latter is relayed via the medial nucleus of the trapezoid body ( M N T B ) . A s depicted in Figure 3.01, the primary projection neurons (principal cells) of the L S O receive a direct glutamatergic excitatory projection from ipsilateral V C N spherical bushy cells (Glendenning, Hutson, et al., 1985; Cant, and Casseday, 1986; Caspary, and Faingold, 1989; Helfert, Juiz, et al., 1992). From contralateral V C N globular bushy cells, an excitatory glutamatergic crosses midline to terminate in the M N T B (Tolbert, andMorest , 1982a; Tolbert, andMorest , 1982b). A glycinergic tract originating from the large spherical M N T B principal neurons then provides inhibition to the L S O (Moore, and Caspary, 1983; Bledsoe, Snead, et al., 1990; Spangler, Warr, and Henkel, 1985; Helfert, Juiz, et al., 1992; Wenthold, Huie, et al., 1987). The two inputs to the L S O are tonotopically matched (Tsuchitani, 1969; Boudreau and Tsuchitani, 1968; Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972; Caird, andKlinke, 1983). Principal neurons of the L S O project primarily to the ICc bilaterally. The ipsilateral projection is inhibitory and glycinergic, while contralateral outputs are glutamatergic or aspartatergic (excitatory). There is an additional excitatory bilateral projection, but it is not well delineated as yet (Saint Marie , Ostapoff, et al., 1989; Glendenning, and Masterton, 1983). 4.3.2 Presynaptic specializations of olivary input neurons The bilateral input neurons o f the superior olive ( V C N spherical and globular bushy cells, as well as M N T B principal cells) possess conspicuous anatomical specializations supporting the representation o f the stimulus waveform. Salient adaptations pertain to the distinctive fibers and synaptic interactions evident in V C N bushy neurons and M N T B principal cells innervating the superior olive. These T. J. A D A M 8 specializations maintain the temporal structure of the signal as it is transferred through ascending pathways to the superior olive, maximizing the fidelity of sound codes. A l l o f the inputs to the M S O and L S O ( A V C N spherical and globular bushy cells, and M N T B principal neurons) possess large diameter axons (Brown, 1987a; Tolbert, and Morest, 1982b). Indeed, the globular bushy cell axons have the largest fiber diameters in the trapezoid body (Tolbert, and Morest, 1982b). Large axonal diameters promote conduction speed by effectively decreasing the longitudinal resistance in the axon. This increases the conduction speed of the action potential (Johnston, 1995; Hi l le , 1992; Matthews, 1986). Due in part to these large axons, the conduction delays are amongst the shortest in the central nervous system. The large diameter axons of these neurons are also heavily myelinated. In the auditory nerve, even the cell somata are myelinated (Liberman, 1980). The resulting insulation increases the velocity o f axonal signal transfer by dramatically increasing the resistance across the neuronal membrane, reducing current shunts (Johnston, 1995; Hil le, 1992; Matthews, 1986). Due to vastly increased conduction speeds, temporal precision in signal transfer across neuron populations is strongly promoted. Even the fine temporal structure of lower frequency sounds (e.g. phase locking up to about 1 kHz) , can be preserved by virtue of the large diameter, myelinated axons (Johnson, 1980; Kiang, Pfeiffer, et a l , 1965; Kiang, Morest, et al., 1973). In the V C N and M N T B , connections between primary projection neurons are characterized by large axosomatic endings, which envelope the majority of the soma o f the recipient neuron, forming a basket-like (calyceal) ending. These specializations are termed the endbulbs of Held in the V C N (Rouiller, Cronin-Schreiber, et al., 1986; Ryugo, and Sento, 1991; Ryugo, and Fekete, 1982; Brawer and Morest, 1975; Tolbert, and Morest , 1982a; Tolbert, and Morest , 1982b), and the calyces of Held in the M N T B (Morest, 1968; Nakajima, 1971). The calyceal endings have multiple synaptic contacts with the recipient neuron. Together, these features promote a fast and uniform polarization of the cell soma by a spike induced post-synaptic potential. The calyceal synaptic contact T. J. A D A M 9 thereby provides a secure and very rapid form o f intercellular communication, supporting temporal fidelity of signal transfer. The shape of the dendritic arbors with which these endings interact supports temporal precision as well. Bushy cells receiving calyceal endings, possess extensive dendritic arborizations, but the calyceal contact is somatic, between the dendrites (Cant, andMorest , 1979a; Cant, 1981; Schwartz, 1977; Henkel, and Brunso-Brechtold, 1990; Lindsey, 1975). The proximity of these large synapses to the spike generating zone o f the postsynaptic neuron reduces the temporal and spatial degradation of the signal imposed by integration over the dendritic tree (Hille, 1992; Johnston, 1995). The auditory brainstem is also specialized by the degree of synaptic convergence onto projection neurons. In A V C N and M N T B , there is minimal synaptic convergence onto the recipient cell, wi th a one to one, or few to one, correspondence of calyceal endings (Ryugo, and Sento, 1991). Although greater convergence is found in the M S O and L S O , this is tonotopically restricted. Generally, input activity is relatively synchronous, minimizing possible timing error in the convergent inputs. A s a consequence, phase-locking is actually enhanced over the auditory nerve, and there is increased security o f the signal transfer to higher centres (Joris, Smith and Y i n , 1994; Joris, Carney, et al., 1994). Together, the described specializations maximize the conduction speed of information to the superior olive, and minimize timing errors. The input fibers, and synaptic interactions in ascending auditory stations innervating the M S O and L S O neurons preserve the temporal fidelity of neural representations o f the sound waveform. 4.3.3 Postsynaptic specializations of olivary input neurons Synaptic interactions at the level of the V C N and M N T B promote temporal precision in signal transfer. Post-synaptic events are large, brief, and exhibit minimal temporal summation and synaptic depression. These features promote the rapid occurrence of brief post-synaptic events, and a quick recovery thereafter. T. J. A D A M 10 In the tissue slice, bushy neurons exhibit brief (2-4 ms) excitatory synaptic potentials (EPSPs). Excitatory postsynaptic currents (EPSCs) are large (40 nS), very rapid (100 (is rise-time, 200 us fall time), and pass through calcium permeable A M P A receptor-coupled membrane channels. These A M P A receptors desensitize rapidly, and the synaptic cleft is cleared quickly of glutamate (Isaacson, and Walmsley, 1996; Zhang, and Trussell, 1994). A s the stimulus strength is raised, EPSCs and resultant EPSPs increase in amplitude, until an action potential is evoked. During repetitive stimulation, EPSPs can follow input activity at rates up to 300 per second, which is comparable to the auditory nerve fibers' maximal firing rate (Oertel, 1985; W u and Oertel, 1984; Oertel, 1983; Wu , and K e l l y , 1991; Isaacson, and Walmsley, 1995; Otis, Zhang, and Trussell, 1996; Otis, Raman, and Trussell, 1995). Together, these features promote the rapid and secure production of large PSPs, and the rapid recovery of the neuron from each synaptic volley, rendering it available for the next one. The V C N bushy neuron and M N T B principal cell are therefore relatively insensitive to previous stimulation, and possess superior temporal tracking ability of the stimulus (i.e. phase-locking). Both the M S O and L S O receive such specialized inputs. Therefore, they are highly specialized to preserve temporal aspects o f the stimulus waveform. 4.3.4 Models of binaural integration Delineation of binaural input patterns to auditory brainstem neurons, and the nature of their interaction at this level have yielded a coherent model for the detection and encoding o f ITDs and IIDs. The frequency dependence o f their effectiveness suggests that IIDs and ITDs are processed independently by the central nervous system. This has been supported experimentally by differential frequency ranges of parallel brainstem pathways processing IIDs and ITDs. The representation o f frequency in the M S O (avian nucleus laminaris) favours the low range (under 4 k H z in small mammals), and ITDs (Guinan, T . J . A D A M 11 Norris, and Guinan, 1972; Goldberg, and Brown, 1968; Parks, and Rubel, 1975). Conversely, high frequencies (> 4 k H z in small mammals) are disproportionately represented in the L S O , which favours the encoding o f IIDs. This correlation has led to the extension o f the duplex theory o f sound to parallel auditory brainstem circuits encoding ITDs and IIDs. 4.3.4.1 The medial superior olive Anatomical and electrophysiological evidence has led to the development of a delay line model o f I T D coding in the mammalian M S O . In this model, interaural disparity coding entails the integration o f binaural excitation in cells that function as coincidence detectors for these bilateral inputs. Through the coding o f synchronous synaptic inputs along the axis of the M S O , this structure produces a spatial code of I T D from phase-locked binaural inputs. From each A V C N , spherical bushy cells exhibit spike discharge that is phase-locked to the stimulus waveform, and provide tonotopically matched excitatory input to the M S O principal neuron (mammal: Rhode, Oertel, and Smith, 1983; Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972; Goldberg, 1969; Goldberg, and Brown, 1968; bird: Parks, and Rubel, 1975; Sullivan, 1985). Synaptic transmission from the ipsilateral and contralateral A V C N s enters the M S O , and courses along input axons that increase in length along the nucleus' axis. Increased lengths of input axons occurs for both bilateral inputs but, according to Jeffress' theory (1948), in opposite directions. This innervation pattern, termed a delay line, introduces lags in the convergence of interaural information in the M S O (Yin , and Chan, 1990; Smith, Joris, et al., 1993; Knudsen, and Konishi , 1979; Konishi , Takahashi, et a l , 1988; Moiseff, and Konishi , 1981; Moiseff, and Konishi , 1983; Sullivan, and Konishi , 1984). The principal neuron of the mammalian M S O , or avian nucleus laminaris, capitalizes on this phase-locking information to measure and encode ongoing temporal T. J. A D A M 12 disparities in the binaural stimulus through the coincidence detection o f binaural inputs over the delay line (dog: Goldberg, 1969; Goldberg, and Brown, 1968; bird: Moiseff, and Konishi , 1983; Moiseff, and Konishi , 1981; Knudsen, and Konishi , 1978; Knudsen, and Konishi , 1979). Interaural time disparities in the binaural signal are offset by a sub-population of M S O neurons occupying a position along the delay line which compensates exactly for the interaural disparity. A t this point, binaural signals collide to produce simultaneous excitatory input to the M S O (laminaris) neuron, which fires maximally in response to the coincidence (Jeffress, 1948; Snyder, 1984). The temporal discharge pattern o f M S O principal neurons is consequently also phase-locked to the stimulus waveform. (Konishi, Takahashi, et a l , 1988; Y i n , and Chan, 1988; Y i n , and Chan, 1990). Therefore, the neural computation performed in the M S O constitutes an information transform from a temporal code in A V C N neurons (phase-locking) to a spatial code in M S O neurons. The I T D of a binaural stimulus is represented by the location o f the M S O principal neuron along the delay line that receives synchronous binaural excitatory input activity at a characteristic delay. Neurons o f the M S O perform an ongoing cross-correlation in the binaural signal, producing a spatial code of I T D within the spatial code o f frequency. 4.3.4.2 The lateral superior olive .v Anatomical and electrophysiological evidence gathered in the mammalian L S O has led to the development of a linear integration scheme for the coding of IID. In this model, interaural disparity coding involves the interaction o f heterologous binaural input information (ipsilateral excitation and contralateral inhibition) i n neurons that function as linear integrators. Through the simple summation o f these inputs in each L S O principal neuron, a rate code of IID is created from phase-locked binaural inputs. T. J. A D A M 13 Tonotopically matched inputs from the ipsilateral A V C N and contralateral A V C N , via the M N T B , exhibit firing that is phase-locked to the sinusoidal stimulus waveform (Young et a l , 1988a; Young et al., 1988b; Rhode, Oertel, and Smith, 1983; Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972; Goldberg, 1969; Goldberg, and Brown, 1968). Ipsilateral excitatory synapses contact, preferentially, the distal portions of the dendrites, while contralateral inhibition arrives at the proximal dendrites and soma (Cant, 1984; Glendenning, Hutson, et al., 1985). This synaptic arrangement is thought to compensate for delays in the timing of inhibition introduced by the presence o f the M N T B . The linear summation o f excitation and inhibition is presumed to produce a steady polarization o f the L S O principal neuron. It is assumed that steady polarization of the somatic membrane potential results from the extensive summation of many small synaptic potentials over the expansive dendritic arbor. This anatomical arrangement is considered to transform a stream of postsynaptic potentials into a stable depolarization by virtue of extensive spatial and temporal summation of minor inputs (Cant, 1981; Tolbert, and Morest, 1982a; Rhode, Oertel, and Smith, 1983). While direct evidence for this integration scheme is lacking, experimental evidence does support the fundamental idea of excitatory and inhibitory integration to produce repetitive firing. For a positive IID, where the sound source is lateralized to the ipsilateral side, a depolarization of the soma is transformed into repetitive firing once spike threshold is reached. In electrophysiological experiments, the rate of repetitive firing has been found to be directly related to ipsilateral excitation, and inversely related to contralateral inhibition (Inhibitory-Excitatory, IE). It is maximal when the ipsilateral side is stimulated alone, and decreases as the contralateral intensity is raised, usually reaching a minimum when binaural stimulation is approximately balanced (Boudreau and Tsuchitani, 1968; Tsuchitani, and Boudreau, 1966; Tsuchitani, 1988a; Tsuchitani, 1988b; Tsuchitani, and Johnson, 1985; Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972). Binaural stimulation of L S O neurons therefore results in spike discharge whose T. J. A D A M 14 rate depends on the relative intensities o f the stimuli at the two ears. This is the strongest evidence supporting integration via the simple summation o f heterologous binaural synaptic drive. Therefore, anatomical and physiological data do indirectly support the summative integration of excitation and inhibition to generate a rate code of IID. A temporal code (phase-locking) is certainly transduced to repetitive firing in the L S O neuron, whose rate may represent IID. Over the entire nucleus, all L S O neurons sensitive to the spectral content of a binaural stimulus presumably respond with a discharge rate, encoding the azimuthal location of its origin. 4.3.5 Temporal response properties of auditory brainstem neurons The temporal firing pattern of auditory neurons has been characterized during brief tone bursts presented monaurally or binaurally to the anaesthetized animal. Typical ly , the stimulus protocol is repeated a number o f times, and the temporal patterns characterized through the formulation o f peristimulus time histograms (PSTHs). These representations constitute ensemble averages of spike latencies during the stimulus (see Methods), and their overall shape describes the temporal discharge pattern o f the cell. 4.3.5.1 V C N inputs to the superior olive Auditory nerve fibers terminating on the spherical and globular bushy cells in the V C N have a distinctive temporal discharge pattern termed the primary response. A t the cell's characteristic frequency (CF), the firing rate is initially high, and gradually declines to a steady state, which persists for the remainder o f stimulus (Ryugo, and Rouiller, 1988; Rhode, Oertel, and Smith, 1983; Romand, 1978). The spherical and globular bushy cells of the V C N receiving the endbulbs of Held also display very high firing rates that decline to a plateau level. Consequently, their firing T. J. A D A M 15 is referred to as the primary-like response (Rouiller, and Ryugo, 1984; Rhode, Oertel, and Smith, 1983; Sento, and Ryugo, 1989). A sub-population of bushy cells exhibits a brief decrease in the probability o f firing immediately after the initial peak in discharge. This is reflected by a notch in the P S T H , and is termed the primary-like-notch response (Rhode, Oertel, and Smith, 1983; Romand, 1978). Principal neurons of the M N T B exhibit primary like spike discharge patterns (Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972; Tsuchitani, 1978; Goldberg, and Brown, 1968), or primary-like-notch behaviour (Smith, Joris, et al., 1989). Therefore, all input neurons innervating binaural cells exhibit similar spike discharge patterns, with rapid firing rates that rapidly decline to a plateau. Transient onset responses are the most prominent response feature. This characteristic promotes fidelity of the auditory signal by introducing minimal spike timing jitter to the auditory code. Transient responses enhance the ability to phase-lock to low frequencies. 4.3.5.2 Principal neurons o f the L S O The temporal discharge pattern o f the L S O principal neuron is termed the chopper response. This is characterized by: (i) a short and very precise onset spike latency, (ii) transient repetitive firing, (iii) highly regular spike discharge that is relatively invariable for a given stimulus intensity, and (iv) spike rate adaptation (Pfeiffer, 1966; Tsuchitani, 1988; Tsuchitani, 1988; Tsuchitani, and Johnson,, 1985; Tsuchitani, 1982; Tsuchitani, 1977; Tsuchitani, and Boudreau, 1966; Boudreau and Tsuchitani, 1968; Guinan, Guinan, and Norris, 1972; Guinan, Norris, and Guinan, 1972; Caird, and Kl inke , 1983). Chopper P S T H s show a number of distinct, regularly spaced modes, which are not related to the stimulus waveform. Instead, they occur as a consequence o f the strict time-locking of discharge onset to that of the stimulus. The multiple peaks apparently occur as a byproduct of the cell's apparently innate tendency to fire at invariant interspike intervals T. J. A D A M 16 (Tsuchitani, 1982; Tsuchitani, and Johnson, 1985; Tsuchitani, 1988; Tsuchitani, 1988; Boudreau and Tsuchitani, 1968). The precisely timed action potentials are the most remarkable feature o f the chopper response, distinguishing it from primary-like discharge, and from simple repetitive firing, which both show variability in spike number and interspike interval duration from trial to trial. These response patterns are therefore irregular. Unlike these forms of firing, the chopper spike count and interspike intervals are highly reliable over repeated presentations of the same tone-burst. The chopper response is distinct, in that it possess both stable spike timing (regularity) and a precise onset latency o f spike discharge from stimulus onset. The chopper response is certainly a unique feature of the L S O principal neuron. The discharge rate is thought to encode IIDs in sound. However, the functional relevance of the precise onset, and subsequent discharge regularity is unclear. Presumably, the onset spike is important for the detection and coding of stimulus onset, but the functional relevance of spike timing precision in chopper responses remains elusive. 4.3.6 Intrinsic membrane properties of auditory brainstem neurons Previous investigations o f the intrinsic electrophysiology o f auditory neurons in the tissue slice preparation have illuminated several key issues regarding response generation at each station o f the auditory brainstem. M o s t of this research has thus far been conducted on the phase-locking and chopper neurons of the V C N , although a preliminary characterization o f chopper response generation in the L S O is available. Because o f the electrophysiological and functional similarities between V C N and L S O chopper neurons, these species are generally presumed to be analogous. In contrast, the primary-like responses of the V C N and M N T B differ dramatically from chopper behaviour in either the L S O or V C N . T. J. A D A M 17 4.3.6.1 Phase-locking neurons of the auditory brainstem Globular and spherical bushy cells of the V C N have nonlinear current-voltage relations. Demonstrated membrane non-linearities result in an emphasis of response onset, and damping responses to polarizations. During suprathreshold depolarizations, only a single action potential is elicited at the onset of the suprathreshold voltage response, over a wide range of current amplitudes (Oertel, 1985; Wu , and Oertel, 1984; Oertel, 1983; Reyes, Rubel, and Spain, 1994). Principal neurons o f the M N T B also exhibit nonlinear current-voltage relations, and elicit a single action potential when spike threshold is crossed (Wu, and K e l l y , 1993; W u , and K e l l y , 1991; Banks, and Smith, 1992). Intrinsic membrane conductances interact in phase-locking neurons to produce low input resistances and rapid membrane time constants. In the depolarized voltage range, low threshold A- type potassium conductances ( G A - revealed through blockade of the underlying channel by 4-aminopyridine, 4 A P ) decrease input resistance. In the hyperpolarizing voltage range, anomalous rectifiers (GKIR> a n d G H - revealed by barium and cesium sensitivity, respectively) perform a similar task (Schwarz, and Pui l , 1998; Banks and Smith, 1992; Borst, Helmchen and Sakman, 1995; Forsythe and Barnes-Davies, 1993; Travagli and Gi l l i s , 1994). Similar membrane properties and conductances have been delineated in V C N octopus cells, which provide phase-locked input to the ventral nucleus o f the lateral lemniscus (Golding, Ferragamo and Oertel, 1998; Golding, Robertson and Oertel, 1995). These intrinsic conductances endow the phase-locker with rapid rise/fall times o f the membrane potential in response to direct depolarization. Additionally, they impede voltage diversions from rest, and prevent repetitive firing. In the listening animal, the distinctive repertoire of intrinsic membrane conductances are thought to promote fidelity in signal transfer through the electrophysiological emphasis o f response onset. A n incoming synaptic volley results in the brief depolarization of the neuron, with/without a single action potential, and a rapid T. J. A D A M 18 resetting o f the neuron subsequent to the transient signal. These properties endow the cells with an ability to follow rapid stimulation rates, with minimal spatial and temporal summation of synaptic potentials, and promote accurate and precise phase-locking to the stimulus waveform. 4.3.6.2 Chopper neurons of the auditory brainstem M o s t o f the research conducted on the intrinsic properties o f chopper neurons has been performed on V C N stellate neurons, which exhibit similar discharge patterns to the L S O chopper neuron, and encode sound intensity through the integration of ipsilateral excitation (Rhode, Oertel, and Smith, 1983; Rouiller, and Ryugo, 1984; Young, Robert, and Shofner, 1988; Young, Shofner, et al., 1988; Hewitt, and Meddis , 1993). However, tissue slice studies performed thus far in the L S O support analogous adaptations for sound coding in the V C N . Chopper neurons possess linear current-voltage relations and long time constants. Min imal membrane rectification results in membrane polarization that directly reflects the magnitude of the stimulus. During suprathreshold depolarization, chopper neurons evoked repetitive firing, the rate of which is proportional to the magnitude o f the depolarization (LSO: W u , and K e l l y , 1993; W u , and K e l l y , 1991; V C N : W u , and Oertel, 1984; Oertel, 1983; Rhode, Oertel and Smith, 1983; Oertel, 1985). In both the V C N and the L S O , chopper neurons are thought to be specialized for the integration of incoming synaptic volleys both in space and time, to produce an output that reflects a time integral of all inputs. Membrane linearity is considered to be the hallmark of this simple summator. While delineation of intrinsic membrane conductances has not been performed in great detail, the absence of obvious membrane rectifications has reinforced the assumption that chopper neurons possess only a minimal repertoire, particularly in the subthreshold voltage range. Instead, the intrinsic membrane behaviour o f the L S O chopper neuron has T. J . A D A M 19 been interpreted as being specialized for the spatial and temporal summation o f synaptic potentials. In turn, membrane linearity of both L S O and V C N chopper neurons has led to the presumption that these neurons function as simple summative comparators of binaural input. In the case o f V C N chopper neurons, convergent excitation leads to depolarization of the somatic membrane, and repetitive discharge whose rate encodes ipsilateral stimulus intensity. In the case o f L S O chopper neurons, convergent excitation and inhibition lead to depolarization o f the somatic membrane, and repetitive firing whose rate encodes IID. The pronounced differences between neurons exhibiting primary-like and chopper firing patterns in the anaesthetized preparation are therefore replicated in the tissue slice. Regular repetitive firing has been attributed to direct summation of synaptic inputs over a membrane exhibiting a long time constant and a linear current-voltage relation. Conversely, nonlinear current-voltage relations, low input resistances, and short membrane time constants, have been considered to be an adaptation for the preservation of temporal input patterns in bushy and M N T B neurons. 4.4 Implications of specializations for the LSO rate code of IID The linear integration hypothesis assumes that L S O neurons integrate ipsilateral excitation and contralateral inhibition to produce chopper output whose rate encodes IID. Further, the unique features of the chopper response are said to arise from this integration. That is, the interplay between convergent excitatory and inhibitory synaptic volleys and the non-rectifying L S O neuron membrane support regular repetitive discharge whose rate is determined by depolarization magnitude at the spike generating zone. However, several other aspects o f the L S O chopper response pattern do not reconcile easily with purported integration mechanisms. T. J. A D A M 20 Hypothetically, integration of many small synaptic inputs over an extensive dendritic arbour produces a steady somatic depolarization which, in turn, supports regular repetitive firing. Additionally, the chopper neuron possesses long membrane time constants which are also thought to contribute to temporal summation yielding regular firing. The relatively minor influence o f each synaptic input minimizes errant spike occurrences and timing errors. This has been assumed to also promote the exquisite timing precision o f chopper onset. However, the linear integration scheme requires that convergent synaptic volleys are integrated over time. The presence o f the early and precise onset spike argues against such a mechanism. The influence o f inhibition on chopper discharge is also emerging to contradict a linear summation model. During monaural stimulation, chopper response synchronization with the stimulus onset decreases gradually. However, binaural stimulation, more adequately resembling the free-field situation, results in decreased discharge rates for sustained discharge, and an abrupt decline o f synchronization in chopper spike trains. Both o f these occur preferentially during later portions of chopper responses. Wi th increased contralateral stimulus intensities, progressively earlier portions o f repetitive firing are affected (Tsuchitani, 1988; Tsuchitani, 1988; Tsuchitani, and Johnson, 1985; Tsuchitani, 1982). The onset spike is the most robust feature of the chopper response, being affected only by very strong inhibition (Tsuchitani, 1988; Tsuchitani, 1988). Previously, these effects were explained via the delayed arrival of inhibition, relative to excitation. This could perhaps be consistent with the linear model. However, intracellular recordings o f L S O neurons indicate that inhibition can precede excitation (Finlay son and Caspary, 1993). Therefore, the integration of excitation and inhibition in the L S O principal neuron is likely more complex and interesting than previously considered. This review prompts the hypothesis that the chopper response is less a feature of the summation of synaptic drive in bilateral inputs than an intrinsic property of the L S O neuron. Synaptic volleys may serve as a trigger for intrinsic membrane properties whose interplay produces the precisely timed onset and regular repetitive firing so unique to T. J. A D A M 21 chopper responses. The specializations o f L S O inputs for preserving the temporal structure o f sound may secure their synchronous arrival at the L S O soma for integration with intrinsic membrane properties. 4.5 The lateral olivocochlear system In addition to encoding interaural disparities for acoustic figure/ground discrimination, the auditory brainstem also participates in the centrifugal control of the organ o f Corti and ventral cochlear nucleus ( V C N ) via the olivocochlear bundle (Warr, and Guinan, 1979; Guinan, Warr, and Norris, 1983). While a paucity o f direct evidence exists, it has been proposed that the descending efferents of the olivocochlear system may play a critical role in signal/noise discrimination by modulating the peripheral auditory system. 4.5.1 LOC cytoarchitecture and circuitry In rodents, the lateral olivocochlear system ( L O C ) originates in the lateral portions o f the SOC; primarily (> 90 %) in the L S O (Aschoff, and Ostwald, 1988). Constituent stellate neurons are small and have a spherical soma, with scanty cytoplasm. Consequently, they are easily overlooked in Niss l stained sections. Dendritic arborizations are slender, branched, and bipolar; extending over considerable distances within the L S O (Helfert, and Schwartz, 1987a; Vetter, and Mugnaini, 1992; Warr, 1975; v Adams, 1983; Brown, 1987b; Liberman, and Brown, 1986; Osen, and Roth, 1969). In small rodents (myomorpha), the somata are dispersed throughout the main body of the L S O , amongst the principal neurons (Osen, Mugnaini, et al., 1984; White, and Warr, 1983; Aschoff, Mul ler , and Ott, 1988; Helfert, Schwartz, and Ryan, 1988; Ryan, Schwartz, et al., 1987). In contrast, L O C neurons of carnivores are concentrated marginally, in each L S O hilus (Warr, 1975; Adams, 1983; L u , Schweitzer, et al., 1987). T. J. A D A M 22 Neurons o f the L O C system receive their primary afferent input (albeit indirectly) from the ipsilateral cochlear inner hair cells. This input arrives via the multipolar neurons o f the ipsilateral posteroventral cochlear nucleus ( P V C N ) , comprising the main output terminus for this structure (Robertson, and Winter, 1988; Thompson, and Thompson, 1991; Warr, 1969). Efferents of the P V C N do not apparently terminate on any other type of L S O neuron (Thompson, and Thompson, 1991). In return, the thin un-myelinated fibers of the L O C projection course mostly ipsilaterally to terminate below the inner hair cells of the same cochlea from which it received its input, primarily on the dendrites of ipsilateral type I spiral ganglion cells (Warr, 1975; Guinan, Warr, and Norris, 1983; Osen, Mugnaini, et al., 1984; White, and Warr, 1983; Robertson, and Gummer, 1985; Robertson, 1984; Strutz, and Bielenberg, 1984; Strutz, and Spatz, 1980; Liberman, 1980). Additionally, collateral projections terminate in the V C N en passant (Spangler, Cant, et al., 1987; Brown, 1988; Ryan, Keithley, et al., 1990). The L O C system therefore forms an intimate feedback loop with the sound transduction mechanism in the ipsilateral cochlea. This appears to be tonotopically organized: The somata o f L O C neurons are distributed across the entire frequency axis o f the L S O , and project tonotopically to the entire length o f the cochlea (Guinan, Warr, and Norris, 1983; Guinan, Warr, and Norris, 1984; Guinan, Norris, and Guinan, 1972). The output neurochemistry o f L O C neurons is complex several transmitter types likely coexisting in synaptic terminals. Immunoreactivity for choline acetyltransferase ( C h A T ) and acetylcholinesterase ( A C h E ) suggest that the descending efferents are cholinergic (Altschuler, Fex, et al., 1984; Osen, Mugnaini, et al., 1984; Osen, and Roth, 1969; Schwarz, Satoh, et al., 1986). However, G A B A , enkephalin, dynorphin, and calcitonin gene-related peptide are also present (Altschuler, Fex, et al., 1984; Vetter, Adams, and Mugnaini, 1991; Abou-Madi , Pontarotti, et al., 1987; Altschuler, Parakkal, and Fex, 1983; L u , Schweitzer, et al., 1987; Schweitzer, L u , et al., 1985; Fex, and Altschuler, 1984; Vetter, Adams, and Mugnaini, 1991; Schwarz, Schwarz, et al., 1988; Schwarz, Schwarz, andHu, 1988). While immunoreactivity studies certainly suggest that T. J . A D A M 23 the L O C projection of the L S O and surrounding periolivary structures co-express a number o f neuromodulator substances, the sources of these inputs remain contentious (Fex, and Altschuler, 1986; Abou-Madi , Pontarotti, et al., 1987; Vetter, Adams, and Mugnaini, 1991). There probably exist several classes of cholinergic and G A B A e r g i c L O C neuron, each co-expressing multiple neuromodulators (Felix, and Ehrenberger, 1992; Vetter, Adams, and Mugnaini, 1991). 4.5.2 LOC physiology andfunction While the anatomy and neurochemistry of the L O C system leads the investigator to certain speculations regarding the physiological function o f the L O C neurons, the exact role of the L O C system remains elusive. It has been proposed that L O C neurons may have a dual purpose: (i) a gain control function in noise protection of peripheral auditory structures from overexposure during high intensity stimulation, and (ii) signal detection in noise. Either suggestion is supported by modulation of electrical events in the cochlea, but underlying mechanisms are as yet unknown (Klinke, and Galley, 1974; Mountain, 1980; Siegel, and K i m , 1982; Masterton, Granger, and Glendenning, 1994). A role in the V C N also evades conclusive discussion. O f the few direct investigations that exist, a study entailing the selective partial ablation of the ventral acoustic stria (i.e. trapezoid body) demonstrated profound deficits in signal (tone) detection in background noise, without compromise o f hearing thresholds for that tone presented in isolation (Masterton, Granger, and Glendenning, 1994). This supports a monaural (presumably L O C ) mechanism for detecting signal tones embedded in noise. Additional indirect evidence stems from the observations that L O C neurons possess fine un-myelinated axons, and employ a number o f neurotransmitters and neuromodulators affecting cochlear function. This suggests a complex control mechanism of inner hair cell transduction, whereby the multiple modulators of the lateral olivocochlear efferents (particularly enkephalins, dynorphins, acetylcholine, and G A B A ) T. J. ADAM 24 exert antagonistic actions on the inner hair cell (Sahley, Kalish, et al., 1991). Their competitive interplay may yield a mechanism for gain control in the protection of auditory structures from overexposure to acoustic stimulation, or provide a signal discrimination mechanism operating peripherally. In light of the foregoing considerations, the overall general function of the olivocochlear system could pertain to improving the signal-noise ratio. The LOC system may modulate the transduction of acoustic stimuli at the inner hair cell level. However, the precise functional effect of this response pattern remains obscure, as does the contribution of the LOC system to auditory function. 5. RESEARCH OBJECTIVES The general purpose of the research reviewed in this dissertation is to determine the validity of the rate code model of LSO output. Experiments focus first on the behaviour of LSO chopper response during prolonged stimulation, as would be expected under natural listening conditions. The report then shifts to delineate chopper response generation by mechanisms intrinsic to the LSO principal neuron. The essence of this research sought to illuminate aspects of the chopper response that might actually be utilized by the central nervous system to encode sound, and determine the mechanisms underlying the generation of this remarkable response pattern. L In the natural listening situation, auditory processing occurs on an ongoing basis. If a rate code of IID is critical for high-frequency sound localization and signal extraction from noise, the chopper rate should remain stable during prolonged stimulation, for a stationary sound source. If so, changes in discharge rate could then be utilized to track a moving sound source. If not, altered discharge rates could reflect either a persistent stationary sound source, or a moving one. This would compromise the utility of the IID code at the level of the LSO. EL The relevance of a short and precise onset latency of the chopper response remains to be delineated. Presumably, it is important for the coding of stimulus onset. If T. J. A D A M 25 so, the onset latency of chopper responses should be preserved following prolonged stimulation o f the L S O principal neuron. A n y shift in onset timing could compromise the detection and coding of a subsequent stimulus. DDL The chopper response pattern has been difficult to relate to the stimulus waveform, and is qualitatively different from either of its primary inputs. It was hypothesized that this rigid temporal pattern is generated, in essence, by the intrinsic membrane properties of the L S O principal neuron. IV. Intrinsic membrane properties are likely endowed to the L S O chopper neuron by underlying membrane conductances similar to those delineated in other neural systems. It was hypothesized that pharmacological delineation o f these conductances would reveal a repertoire of intrinsic determinants of chopper behaviour. These conductances would dictate the modulation o f chopper responses by membrane potential. V. The integration mechanism and relative timing of convergent excitation and inhibition is unresolved. We sought to illuminate this issue by examining the voltage dependence of chopper firing. O f course, the increased membrane conductance associated with the activation o f glycine-activated chloride channels is not accounted for in this experiment. However, the voltage dependence o f intrinsic membrane properties must mediate the interaction between excitation and inhibition, and therefore contribute to their integration. VL Finally, the intrinsic membrane properties of the L S O principal neurons were hypothesized to play a role in chopper modulation during prolonged stimulation. Given that chopper expression is influenced critically by the intrinsic electrophysiological properties of the membrane, these may also participate in chopper response modulation during prolonged stimulation. T. J. ADAM 26 6. E X A M I N A T I O N OF S H O R T - T E R M A D A P T A T I O N IN T H E SUPERIOR O L I V E Introduction It is well understood that the azimuthal location of sound sources, and the acoustic shadow of the listener's head and shoulders, impose interaural intensity disparities (IIDs). We assume that these IIDs are encoded as the discharge rate of IE responses in LSO neurons. Discharge rate may provide, therefore, a code that enables higher auditory centres to distinguish multiple sound sources by their location along the azimuth. Over the spectral map of the LSO, the different frequency components of a complex stimulus may produce comparable discharge rates in IE neurons. Matches in discharge rate would then provide a basis for the perception of the complex stimulus as a unique sound at a specified location. Principal neurons of the LSO receive excitatory input that originates in the ipsilateral ear, and inhibitory input that originates in the contralateral ear. These inputs, combined in the principal cell, determine the output firing rate. For sound sources located on the ipsilateral side, excitation exceeds inhibition, and the net depolarizing current presumably regulates discharge rate. For each ipsilateral stimulus location then, we presume that converging excitation and inhibition produce a degree of depolarization which translates into the rate of repetitive action potential discharge. The rate of repetitive firing reflects, therefore, the balance between excitation and inhibition imposed by sound source location. The rate code model predicts the stability of IE neuron discharge rate over the stimulus period, even for sounds that vary in intensity over time. Since rate changes would indicate movement of the sound source, spike discharge rate must remain stable for stationary sound sources. Little is known about the required stability in the firing rate output because previous electrophysiological studies have typically employed tonal stimuli of short duration to evoke spike discharge. In order to assess the validity of the T. J. A D A M 27 rate code hypothesis, we need to know i f the discharge rate for a given IID is preserved during prolonged stimulation. The prevailing hypothesis is that responses of EE L S O neurons represent a stable rate code o f IID for non-moving sound sources throughout the stimulus period. Therefore, responses in the L S O should, ideally, exhibit no changes in responsiveness as long as the sound source does not move. This hypothetical requirement is difficult to reconcile with any kind o f adaptation. A shift in discharge rate during a sound stimulus would introduce ambiguities in the putative rate code for IID. This would indicate that the firing rate of individual L S O neurons cannot adequately represent sound direction. The latency of response onset is another aspect of discharge behaviour worth examining in EE L S O neurons. Although it does not appear to relate systematically to IID, it may play a role in the detection and tracking o f stimulus changes over time. Timing of intensity fluctuations in an ongoing stimulus, or o f a stimulus occurrence during repetitive stimulation would be encoded by the stimulus-locked onset response. The onset latency of IE neuronal discharge should also, therefore, remain stable over time during repetitive stimulation. The following series o f experiments examines short-term adaptation in binaural EE neurons of the L S O . Prolonged stimulation o f each (ipsilateral and contralateral) input is achieved through dichotic stimulation using tones. Excitatory and inhibitory interactions are then examined during binaural stimulation. For each experiment, the discharge rate and onset latency are measured to evaluate possible short-term adaptation effects on neuronal responses. Experiments are presented in the following order, (i) Spike discharge is examined during short ipsilateral (excitatory) and binaural stimulation, over a range of stimulus intensities. This phase outlines the interaction of excitation and inhibition to determine the EE response elicited during binaural stimulation, (ii) Short-term adaptation is examined during prolonged excitation, (iii) Short-term adaptation in inhibition is examined during prolonged stimulation, (iv) Binaural responses during prolonged stimulation are studied T. J. A D A M 28 with equal interaural intensities (IID = 0). (v) Binaural responses during prolonged stimulation are examined with unequal interaural intensities (IID > 0). (vi) Post-stimulatory changes in discharge rate are appraised for each stimulus configuration, (vii) Post-stimulatory changes in onset latency are evaluated for each stimulus configuration. 6.1 Methods 6. J. 1 Selection of animals The investigation of short-term adaptation in neurons of the superior olivary complex (SOC) was conducted in young adult Long Evans rats (3 to 6 months o f age) weighing 400 to 700 g. This species is ideal for the study of L S O electrophysiology because their L S O is large, and their audiograms indicate superior high frequency hearing (Masterton, Granger and Glendenning, 1994; Harrison and Howe, 1974). The pigmented Long Evans strain does not exhibit the albino-related hearing loss of Sprague Dawley and F-344 rats (Finlayson and Caspary, 1993; Finlayson, 1994). Accordingly, sufficiently sensitive hearing could be demonstrated during the measurement of auditory brainstem evoked responses. 6.1.2 Surgical preparation for acute electrophysiological experiments Animals were anaesthetized with sodium pentobarbital (44 mg/kg, intra-peritoneal injection), xylazine (3 mg/kg), and ketamine (85 mg/kg; intra-muscular injection). Supplemental doses of the anaesthetics were administered as required, to prevent the observation o f eye-blink and toe-pinch reflexes (every 2 to 3 hours). Sodium pentobarbital and ketamine supplements were given in alternation. The initial surgery entailed dissection to expose the tympanic membrane for access by the acoustic stimulation couplers. The ectotympanic rim and external auditory canal were exposed by retracting the pinnae and the external auditory meatus. The tympanic membrane was inspected for pathology using a dissection microscope. Only T. J . A D A M 29 animals free o f effusion or cerumen were included in this study. In addition, baseline auditory thresholds were determined for each ear using monaural clicks and recording auditory brainstem evoked potentials ( A B R ) prior to each recording session. This ensured the adequacy of hearing sensitivity in animals studied. Animals were fixed in the standard stereotaxic plane with a head holder fixed to the dorsal cranium using jeweler's screws and cyano-acrylate glue. Ho l low stereotaxic ear bars attached to stimulus transducers were placed against the tympanic rim to form a closed stimulus delivery system. Standing waves were minhnized using cotton and steel wool placed in the coupling tubes. Neck musculature was reflected from the posterior surface of the skull, to allow access to the caudal portions o f the skull. The foramen magnum was enlarged dorsally, and the dura mater carefully reflected to allow for electrode access to the L S O through the intact cerebellum. During surgery and the subsequent experiment, the animal's rectal temperature was maintained at 37 °C by a thermostatically controlled D C heating pad. 6.1.3 Stimulus generation and delivery Auditory stimuli were generated digitally, and delivered using the program A/Dvance (McKel la r Designs) on a Macintosh Quadra 950 computer with National Instruments* boards (MIO-161-9; D M A - 2 8 0 0 ) . Output of the digital-to-analog converters was band-pass filtered, amplified (Amcron amplifier, Crown), and attenuated under computer control (Medical Associates M A 9 1 9 attenuators). In the stimulus transduction couplers, stimuli were delivered through impedance-matched headphones (Beyer-Dynamic 600 W , B4-132.01; frequency range 0.05 to 35 kHz) . The output of the stimulus generation- and delivery system was calibrated off-line using a transduction coupler with a volume equal to the external auditory canal (0.3 ml). The amplitude o f the auditory signal was measured with a 0.25 inch condenser microphone (Bruel and Kjaer) and a calibrated microphone preamplifier (Bruel and Kjaer T. J. A D A M 30 2804). Calibration tables were then computed to deliver desired intensities in dB SPL (re: 0.0002 dynes/cm 2) of tonal stimuli generated within the frequency range o f 0.05 to 32 k H z . 6.1.4 Extracellular recordings Recordings from single units in the S O C were achieved using microelectrodes to approach single neurons and record action potentials. These microelectrodes were pulled from borosilicate glass and had outer tip diameters of 0.5 to 1.0 um. Filling the electrode with 2 % wheatgerm agglutinin horseradish peroxidase ( W G A - H R P , for extracellular marking) in 2 M N a C l then produced tip resistances of 10 to 20 MQ. Electrode recordings were amplified using an AM-Sys tems 1800 amplifier and the Kik i su i oscilloscope, and band-pass filtered between 10 H z and 5 k H z . Spikes were identified on the oscilloscope, and converted to digital pulses using a WPI 121 window (voltage threshold, 10 us resolution) discriminator, for subsequent input to the A/Dvance recording module. To reach the L S O , recording electrodes were advanced into the brain at a compound angle o f 49.5 0 to the right o f midsaggital axis, and 15° up from the horizontal axis, to reach the left SOC. A s the recording electrode approached the depth o f the superior olive, search stimuli comprised o f binaural clicks were delivered repetitively (2/second), and evoked potentials were recorded Continuously. Clicks delivered to the contralateral (right) ear were delayed 20 ms from those delivered to the ipsilateral (left) ear. Recordings indicated electrode penetration of the SOC when evoked potentials exhibited two peaks of opposite polarity with latencies between 5 and 20 ms following each click. These potentials became approximately equivalent in amplitude as the L S O was reached. The relative polarity and magnitude o f ipsilaterally and contralaterally evoked potentials proved to be a highly reliable indicator of electrode position near and within the SOC. However, following each experiment, the recording position was T . J . A D A M 31 confirmed through histological examination of W G A - H R P marks at the site of ejection from the electrode tip. Extracellular recording sessions for individual auditory units commenced with an initial characterization o f the frequency response using 50 ms sinusoidal stimuli (5 ms rise/fall). The best frequency (BF) for each unit was measured as the stimulus frequency at which the unit generated the highest spike count. This was determined from ipsilateral and contralateral monaural iso-intensity curves at 50 dB SPL. For binaural units, each ear was tested to confirm that bilateral inputs were tonotopically matched. Thresholds at B F were determined from monaural rate-intensity functions, and inhibitory thresholds from binaural rate-intensity functions. Finally, aurality (the pattern of excitation and inhibition from each ear) was determined on the basis o f responses to binaural B F stimulation at varied interaural intensity levels. Binaural units excited by ipsilateral stimulation, and inhibited by contralateral stimulation (IE pattern), and residing in the L S O , were presumed to be the principal neurons of this nucleus, largely due to the uniqueness of their response characteristics within the auditory brainstem. Monaural units responding with either ipsilateral or contralateral excitation were taken to comprise a mixed group, that included the primary input projection neurons from the cochlear nuclei or medial nucleus of the trapezoid body to the L S O . 6.1.5 Paradigms addressing short-term adaptation Short-term adaptation of S O C units was invoked using several configurations of paired tone stimulation (e.g. Boettcher, Salvi and Saunders, 1987; Boettcher, Salvi and Saunders, 1990; Chimento and Schreiner, 1991; Kaltenbach, Meleca, et al., 1993). The first stimulus of the pair was an adapting stimulus (200 ms B F tone), and the second was a probe stimulus (30 ms B F tone). These tones were presented at equal intensity, 20 to 40 dB above threshold, and had 5 ms rise/fall ramps. The delay between the adapting tone and probe tone was varied to address the time course of short-term adaptation. Probe T. J. A D A M 32 delays were 1,2,4, 8, 16, 32, 64,128, 256, 512 ms from the offset of the adapting tone. For each delay, 30 trials (adapting stimulus - probe stimulus) were conducted in random order o f presentation. A t an inter-trial interval o f one second, carryover effects were not observed. Responses during adapting tones were employed to evaluate the magnitude and onset time course o f short-term adaptation, while the responses evoked by probe tones at various delays provided a measure for the time course of recovery. Short-term adaptation was evaluated in each primary input pathway to the unit being recorded by stimulating the appropriate ear. For monaural units, tone pairs were presented to the ipsilateral or contralateral ear, depending on the unit's aurality. For IE units (presumptive L S O principal neurons), short-term adaptation in the ipsilateral input pathway was invoked through presentation of tone pairs to the ipsilateral ear only. Because inhibition cannot be measured extracellularly in IE neurons in the absence of firing, we measured its effect on ipsilaterally evoked spike discharge. Therefore, short-term adaptation in the contralateral input pathway entailed the presentation o f adapting tones to the contralateral ear, and simultaneous brief (30 ms) tones to the ipsilateral ear at various delays into each adapting stimulus (1, 2, 4, 8, 16, 32, 64, 128 ms). Similarly, recovery of inhibition from short-term adaptation was examined with contralateral adapting tones followed by binaural probe tones. Binaural short-term adaptation was assessed through the presentation o f binaural adapting stimuli followed by binaural probe stimuli. A possible dependence o f short-term adaptation on the intensity of stimulation necessitated the presentation of adapting and probe tone pairs at several levels of stimulus intensity. Short-term adaptation was evaluated over the intensity range from -12 dB SPL (near response threshold) to ~ 60 dB SPL. Additionally, a range (0 to 45 dB SPL) of interaural intensity disparities was employed in the examination o f binaural short-term adaptation. This range covers the physiological range o f IIDs in the rat (Finlayson and Caspary, 1993), and those encoded by L S O neurons in the cat (Irvine, 1987). Significant T. J. A D A M 33 effects of IID and overall stimulus intensity were observed in the short-term adaptation o f binaural units. 6.1.6 Assessment of short-term adaptation Peri-stimulus time histograms (PSTHs, 1 ms binwidths) for adapting and probe tone responses were constructed from spike occurrences during 30 trials to determine the spike latency distribution. This representation was used to quantify responses, and measure short-term adaptation in S O C units. Spike discharge was quantified as the average spike count per bin in each P S T H over 30 trial presentations, integrated over the first 30 ms of the response. The onset latency of spike discharge was then measured for each stimulus. The latency was defined as the time o f the first bin that exceeded the background firing ( i f present) by 2 standard deviations. These indices quantified the response level and the onset time o f evoked discharge. Each of these measures was expressed as mean ± standard error. The size of response decrements during the adapting tone and probe tones reflected the magnitude of short-term adaptation. This was normalized as a percentage o f the fully recovered discharge level. [ Probe response @ 1ms delay Magnitude o f short- term adaptation (%) = 1 • Probe response @ 512 ms delay *100 The time course of short-term adaptation was assessed for each unit as the time constant of an exponential function fit to the response during the adapting tone, and to the magnitudes of probe responses (above) as a function o f the delay from the adapting tone. Both single and double exponential functions were fit to data, but only the former was used to describe short-term adaptation, as double exponentials did not yield significantly superior fits: The formula for the single exponential function used was T. J . A D A M 34 y = k + k e 3 0 1 where the asymptotic response magnitude (steady-state response) was ko, the ratio of steady-state response level to peak response was k l 5 and tau (x) was the time constant. Short-term adaptation was examined in P S T H s constructed from 300 repetitions (1 per second) of each adapting tone. Exponential functions were fit to responses, and the exponential time constants were used to describe the time course of short-term adaptation. The time course of recovery from short-term adaptation was derived from the magnitudes of probe responses. These were plotted as a function o f their delay from the adapting stimulus, and fit with single exponential functions. We examined the interaction o f excitation and inhibition in binaural EE responses during prolonged stimulation. Short-term adaptation and recovery during ipsilateral, contralateral, and binaural stimulation were compared using a one-way repeated measures analysis of variance ( A N O V A ) . Data for stimulus intensities over the entire functional range (above) were included for each unit. 6.1.7 Histological verification of neuron location The location of EE neurons within the L S O was confirmed histologically, following each experiment. Recording sites and tracks were marked by ejecting wheatgerm agglutinin-horseradish peroxidase ( W G A - H R P ) from the electrode following each recording session. Presentation of 5 u A positive current pulses (500 ms duration) at 1 H z for 5 minutes at the recording site marked the position o f the electrode. Subsequently, tracks were marked by ejecting W G A - H R P while slowly retracting the electrode. A t the end of each experiment, the animal was sacrificed by an overdose o f sodium pentobarbital and perfused through the heart with 0.9 % saline followed by 1.0 % paraformaldehyde and 1.2 % gluteraldehyde in 0.1 M phosphate buffer with 10 % sucrose. Frozen sections were cut in the plane o f the electrode track, mounted, and T. J. A D A M 35 reacted for peroxidase (tetramethylbenzidine/glucose oxidase chromagen reaction, K i t a and Armstrong, 1991). Sections were counterstained with thionin. Camera lucida drawings of relevant sections and our records of electrode depth for each recording were used to identify locations of the neurons studied. 6.2 Results Extracellular recordings were conducted on 69 neurons residing within superior olive, 37 of which were in the L S O . Units were classified according to location, aurality, and discharge pattern. For IE units, presumed to be L S O principal neurons, the short-term adaptation patterns in excitation and inhibition were delineated. Their respective influences on the behaviour o f binaural responses during prolonged stimulation were then investigated. Finally, short-term adaptation was examined in monaural units (EO, OE). These units likely include the input neurons o f the L S O from V C N bushy cells and M N T B neurons. Results from this experimental series have been published previously (Finlayson and Adam, 1997; Finlayson and Adam, 1996). 6.2.1 Location of recordings within the superior olive In order to identify neuronal populations o f interest (binaural L S O neurons and their monaural input neurons), the location of each extracellular recording within the SOC was confirmed through the ejection of a marker ( W G A - H R P ) from the electrode. Subsequent histological study revealed the location of each recording site and the track followed by the electrode through the brainstem. The recording site, aurality, and discharge pattern (e.g. chopper response) were used to classify each unit. Figure 4.01 illustrates the localization o f W G A - H R P ejected after each o f three recording sessions in a single electrode penetration. Each recording site is identified by an T. J. A D A M 36 area of increased marker diffusion. In this animal, one unit was located in the trapezoid body ventral to the L S O {filled arrowhead), and two more units recorded in the dorsal flexure of the L S O (arrow). In all, 37 units were recorded in the L S O , out of a total of 69 units confirmed to reside in the SOC. Within the boundaries of the L S O , both monaurally and binaurally driven units were encountered (n=37). Binaural units responded to ipsilateral stimulation with excitation, and contralateral stimulation with inhibition (IE, n=25/25). Localization studies restricted their distribution to the L S O , and therefore support their identity as the primary projection (principal) neurons of the L S O . Monaural units (OE or EO) were observed in all regions of the S O C (n=44), including the L S O (n=12), and exhibited primary-like discharge patterns (n=44). These units were excited by stimulation o f either the ipsilateral ear (OE, n=35) or the contralateral ear (EO, n=9). These units may correspond to the input neurons for the SOC. Histological localizations support the correspondence of binaural IE neurons recorded in these experiments to the principal neurons of the L S O . The monaural units exhibiting primary-like discharge and ipsilateral or contralateral excitation may be a mixed population, the bulk of which probably represents the primary projection neurons from the ventral cochlear nuclei and M N T B to the L S O . Our investigation of short-term adaptation in the S O C is limited to these populations. 6.2.2 Spike discharge of IE units during ipsilateral and contralateral stimulation 6.2.2.1 Responses to ipsilateral excitation Spike discharge was evoked during monaural ipsilateral stimulation. The firing pattern of IE neurons in the L S O was initially identified as the chopper response by the T. J. A D A M 37 superposition of action potential occurrences viewed on a storage oscilloscope. Responses were recorded during 30 presentations of the ipsilateral stimulus, and P S T H s constructed from their latencies. A range o f ipsilateral stimulus intensities was tested to determine the relation o f spike discharge to sound level. This is depicted in Figure 4.02 A for a typical EE unit, where the chopper pattern was masked by the sampling of spike latencies into bins o f relatively long duration (1 ms). Three ipsilateral stimulus intensities yielded these PSTHs . In each, firing began early in the stimulus period, with a short onset latency and relatively high discharge rate ( P S T H peak, arrow). Firing rate subsequently declined towards steady-state within 50 ms ( P S T H plateau, *). Both the steady-state (*) and peak (arrow) discharge levels were directly related to the ipsilateral stimulus intensity. However, peak responses became more prominent with increased stimulus intensity. Thus, peak responses grew with ipsilateral stimulus intensity more than steady-state responses did. 6.2.2.2 Responses to binaural stimulation The influence o f contralateral inhibition on ipsilaterally evoked chopper responses was examined during binaural stimulation of L S O neurons. A range o f contralateral stimulus intensities were tested for a constant ipsilateral level. For the same unit in Figure 4.02 A , binaural spike discharge is shown in Figure 4.02 B . Ac t ion potential firing was evoked by an ipsilateral tone o f intensity 39 dB SPL (B, left). Binaural stimulation with an ipsilateral intensity of 39 dB SPL, and contralateral intensities of 24 and 31 dB SPL yielded the other P S T H s (middle and right, respectively). Inhibition had clear but complex effects on spike discharge. For increasing contralateral intensity, inhibition first decreased steady-state responses (Figure 4.02, middle and right PSTH). A s the inhibitory intensity approached the level o f the ipsilateral stimulus, both sustained and peak discharge were decreased, producing a minimal response. The onset portion o f spike discharge was unaffected by inhibition until T. J . A D A M 38 the contralateral stimulus level was raised to a level sufficient to suppress the response entirely. Therefore, the arrival of inhibition at the L S O neuron decreases later sustained portions of responses preferentially. The latency o f response onsets was also examined for increasing ipsilateral and contralateral stimulus intensities. Response latencies shortened with ipsilateral intensity over a narrow range (6.7 to 3 ms, 10 dB SPL increase), as shown graphically in Figure 4.02 C (left). This was consistent over the sample o f units tested. Onset latencies decreased from 9 ± 2.3 ms to 3.2 ± 0.6 ms over the ipsilateral intensity range o f 12 to 45 dB SPL. Although the range o f onset latencies observed was narrow (~ 5 ms), changes in latency were significantly related to ipsilateral intensity ( A N O V A , p<0.05). Inhibition did not alter the onset latency of spike discharge, until its intensity was sufficient to suppress the responses (Figure 4.02 C , right). Consequently, the ipsilateral stimulus intensity is the primary determinant of chopper onset latency o f binaural L S O units. In sum, the discharge rate and onset latency of IE responses are largely determined by the intensity of the stimulus presented to the ipsilateral ear. The latency of firing is shortened, and discharge rate increases with stimulus intensity. Contralateral inhibition preferentially suppresses steady-state discharge, rendering the early peak discharge the most robust feature o f firing. Thus, the IID code would appear to be better carried in the steady-state firing than in the onset peak o f action potential discharge. In our study of binaural short-term adaptation in L S O principal neurons, therefore, we examined both the discharge onset and average discharge level (spikes/bin) of firing. 6.2.3 Short-term adaptation of excitation and inhibition in binaural LSO units The azimuthal location o f a sound source is directly related to interaural intensity disparities (IIDs) in auditory signals reaching the ears (Shaw, 1974; Phillips, Calford, et a l , 1982). It is commonly assumed that L S O principal neurons encode the IIDs in the T. J. A D A M 39 discharge rate of chopper responses. To prevent ambiguity, the rate code for IID would be expected to remain stable during prolonged stimulation by a non-moving sound source. Chopper discharge rate is determined by the integration of ipsilateral excitation and contralateral inhibition. This integration presumably creates a net depolarization whose magnitude is translated directly into chopper discharge rate. I f L S O neurons were to exhibit no short-term adaptation in these inputs, the stable rate code required by our hypothesis would be guaranteed. However, short-term adaptation has been described previously in the cochlear nucleus containing the primary input neurons of the L S O (Boettcher, et al., 1990; Boettcher, et al., 1987). The adaptation pattern o f L S O afferents could be fed forward to the principal neuron. Therefore, short-term adaptation may be evident in the excitation of L S O neurons, generating ambiguity in the information encoded by firing rate. It is important, therefore, to determine how the L S O neurons transform the adapted input signals into a meaningful output. 6.2.3.1 Ipsilateral excitation Our experiments begin with the examination of chopper responses evoked during prolonged tonal stimulation. Ipsilaterally presented tones of extended duration could induce adapted response patterns in chopper neurons, reflected by decreases in discharge rate during the chopper response. I f so, the magnitude and time course o f changes in responsiveness require characterization, in order to understand which central compensation would be necessary to salvage the rate code for sound direction. Excitatory short-term adaptation was examined in chopper responses to 200 ms B F adapting tones delivered to the ipsilateral ear. Because of the intensity dependence of chopper responses, adapting stimuli were presented at several intensities to assess possible changes in short-term adaptation with stimulus level (12 dB SPL to 45 dB SPL). Changes in responsiveness were then evaluated in P S T H s constructed from spike firing T. J. A D A M 40 sampled during 30 repetitions of each stimulus tone. These histograms characterized, then, any time-dependent changes in spike discharge level (spikes/bin) during the tone. Short-term adaptation was evident in P S T H s as decreases in spike discharge towards a plateau. Figure 4.03 depicts this change in chopper discharge rate for a typical binaural L S O unit. Four P S T H s are shown, over a range of ipsilateral stimulus intensities (29 to 44 dB SPL, Figure 4.03 A ) . Each P S T H shows an early onset response characterized by a high discharge rate (arrow), which decreased to a plateau (*). This is indicative of short-term adaptation in monaural excitation for the IE neuron. Comparing the four P S T H s reveals the effect of ipsilateral stimulus level on L S O principal cell discharge. A s the ipsilateral stimulus intensity was increased, sustained discharge levels (*) increased. Additionally, a more prominent onset response (peak, arrow) emerged in the chopper responses (Figure 4.03 A ) . Therefore, both the peak discharge and sustained discharge were dependent on intensity, however, to different degrees (above). This disparity indicates that short-term adaptation in excitation is dependent on ipsilateral stimulus intensity. Both the magnitude and time course o f short-term adaptation were influenced by stimulus intensity. For our exemplar, the size of response decrements rose from approximately 30 % near response threshold (Figure 4.03 A , bottom PSTH) to 83 % at 44 dB SPL (Figure 4.03 A , top PSTH). This is depicted graphically in Figure 4.03 B , where the magnitude o f adaptation approached an asymptote near 95%. Exponential time constants describing the time course of short-term adaptation became shorter with increasing stimulus intensity, decreasing from 33 ms near response threshold to 3 ms at 44 dB SPL (Figure 4.03 C). Taken together, these indices show that the strength o f short-term adaptation increases with stimulus level. Thus, greater response decrements occur more quickly during more intense ipsilateral stimulation. Across the sample of EE units studied, all exhibited short-term adaptation in excitation during ipsilateral stimulation (n=25/25). The magnitudes and time courses of short-term adaptation were always dependent on stimulus intensity. Raising stimulus T. J. A D A M 41 level significantly increased response decrements, and accelerated short-term adaptation ( A N O V A , p<0.01). In the range of 12 to 45 dB SPL, responses to prolonged ipsilateral stimuli decreased by 30 to 95 %, with time constants of 38 to 3 ms. For comparison to the short-term adaptation o f contralateral inhibitory inputs and to other neuronal responses, averages were calculated for the magnitude and time course of excitatory short-term adaptation across stimulus intensity (response threshold near 12 dB SPL to maximal responses at 45 dB SPL). O n average, the firing rate of L S O neurons decreased by 68 ± 4.4 % of their peak rates. During intense stimulation, the adapted rates of all neurons approached an asymptote near 5 % of the peak rate. The exponential time constants for short-term adaptation averaged 5.11 ± 0.96 ms, and approached 3.5 ± 0.5 ms for the most intense stimuli. This value may approximate the maximal short-term adaptation rate for binaural L S O neurons. These results demonstrate that short-term adaptation does occur in chopper responses evoked during ipsilateral stimulation. O n average, responses are decreased by ~ 70 % along a time course characterized by the exponential time constant o f ~ 5 ms. Furthermore, the intensity o f the ipsilateral stimulus has pronounced effects on short-term adaptation, determining both its magnitude and time course. With more intense stimulation, short-term adaptation becomes greater and faster in IE units. 6.2.3.2 Contralateral inhibition Having demonstrated that short-term adaptation occurs in L S O IE neurons, our attention now turns to short-term adaptation o f inhibition in these units. Excitatory short-term adaptation implies that the hypothesized rate code of IID may shift during prolonged stimulation. Therefore, short-term adaptation may also occur in the inhibition o f L S O neurons. Further, the magnitude and time course o f inhibitory short-term adaptation are hypothesized to match those of excitatory short-term adaptation. The T. J. A D A M 42 equivalence o f short-term adaptation in L S O inputs could represent a compensatory mechanism to preserve the rate code of IID during prolonged binaural stimulation. Short-term adaptation of inhibition was studied in the spike discharge evoked during 30 ms ipsilateral (excitatory) tones presented at various delays from the onset of a 200 ms contralateral (inhibitory) adapting tone. Short-term adaptation in inhibition was produced by the contralateral adapting tone, and measured during ipsilateral tones. Ipsilateral stimulus intensity, sufficiently intense to produce robust spike discharge, was held constant to examine inhibitory adaptation. Changes in IE neuron discharge could then be related directly to short-term adaptation of inhibition. Increased firing rates during 30 ms ipsilateral tones were evidence o f inhibitory short-term adaptation. Robust spike discharge, sufficient to examine inhibitory short-term adaptation, could be evoked in 19 of the 25 binaural units studied. In the remaining 6 only minimal spike discharge could be observed, even at low contralateral intensities. This precluded the examination o f inhibitory adaptation on spike discharge. Therefore, only the 19 IE units exhibiting sufficient spike discharge were included in this analysis. O f these IE units, all exhibited short-term adaptation in inhibition. Short-term adaptation o f inhibition was evident in the P S T H s for a typical L S O IE unit depicted in Figure 4.04 A . Each P S T H depicts the spike discharge during the 30 ms ipsilateral stimulus presented at a particular delay from the onset o f the 200 ms contralateral adapting stimulus. Comparison of P S T H s demonstrates that spike discharge increased with the ipsilateral stimulus delay. Therefore, short-term adaptation of inhibition occurs in L S O neurons. For this unit, the discharge rate (average spikes/bin over 30 ms) increased by 136 % at a contralateral stimulus intensity o f 17 dB SPL (ipsilateral intensity 37 dB SPL). A t this intensity, short-term adaptation occurred with a time constant 7.26 ms. This is clear evidence o f the influence of inhibitory short-term adaptation on ipsilaterally evoked spike discharge in IE neurons. The effect o f stimulus intensity on inhibitory short-term adaptation was examined by manipulating the intensity of the contralateral adapting tone, while holding the T . J . A D A M 43 ipsilateral stimulus level constant. The discharge rate (average spikes/bin) was plotted against inter-stimulus delay at each contralateral intensity tested. For our exemplar unit of Figure 4.04, the magnitudes and time courses of inhibitory short-term adaptation are depicted for two contralateral stimulus intensities (17 and 12 dB SPL) , while the ipsilateral intensity was 37 dB SPL (Figure 4.04 B) . A s illustrated in Figure 4.04 B for a typical unit, lower discharge rates and greater inhibitory short-term adaptation (*) occurred as the contralateral intensity was raised. Additionally, the time course of short-term adaptation became more rapid with greater stimulus intensities (9.52 ms at 12 dB S P L , and 7.26 ms at 17 dB SPL). Over the entire sample of EE neurons, short-term adaptation in inhibition depended on the contralateral stimulus intensity. The magnitudes and time constants of short-term adaptation averaged over stimulus intensity (12 to 35 dB SPL) were 111 ± 30 %, and 6.9 ± 1.1 ms, respectively. Previously, it was noted that inhibition had almost no influence on the onset latency of spike discharge evoked during ipsilateral stimulation (Figure 4.02 C) . This was further supported in this study of inhibitory short-term adaptation. Over the entire range o f contralateral stimulus intensities examined, inhibitory short-term adaptation had no effect on the onset latency of chopper discharge, unless the intensity was sufficient to suppress the entire chopper response (Figure 4.04 D) . Therefore, inhibition bears no influence on the onset timing of chopper responses, unless firing is prevented altogether. In comparison to excitation, the indices of average magnitude and time constant initially appear to reflect greater and slower short-term adaptation in inhibition. Excitation decreased responses by 70 %, with a time constant of 5 ms. Inhibition increased responses by 110 % with a time constant of 7 ms. However, the intensity range employed to investigate inhibitory short-term adaptation was narrower than that that used in excitatory experiments (inhibition: 12 to 35 dB SPL; excitation: 12 to 45 dB SPL). T. J. A D A M 44 Accounting for respective stimulus intensities, the magnitude and time course of inhibition did not differ from those of excitation ( A N O V A , p<0.05). These results demonstrate that short-term adaptation of chopper responses occurs in L S O neurons during contralateral stimulation. Additionally, the onset latency of chopper responses is not influenced by inhibition at any contralateral stimulus intensity. During short-term adaptation, responses are increased by an average 110 % with an average time constant of 7 ms. The intensity o f the contralateral stimulus has pronounced effects on inhibitory short-term adaptation, determining both its magnitude and time course. With more intense stimulation, inhibitory short-term adaptation is greater and faster in EE units. Furthermore, the magnitude and time course o f inhibitory short-term adaptation is not significantly different from those of excitation when stimulus intensities are accounted for. Therefore, during free-field binaural stimulation, short-term adaptation in excitation and inhibition may be equivalent over time, yielding a stable rate code of IID. 6.2.3.3 B inaural responses Short-term adaptation has been demonstrated above in both the excitation and inhibition of EE L S O neurons. The magnitudes and time constants o f short-term adaptation show similar dependence on stimulus intensities for both excitation and inhibition. The possibility, therefore, exists for the preservation of the rate code o f IED in L S O neurons through a balance o f excitatory and inhibitory short-term adaptation. In the simplest scenario, hypothetical short-term adaptation of excitation and inhibition would be o f equal magnitude and time course during binaural stimulation o f the L S O neuron. This would avoid an ambiguity in IID information encoded by firing rate at different times of the response to prolonged sound. Through short-term adaptation of inhibition, L S O neurons could therefore transform adapted excitatory input signals into a stable output. T. J. A D A M 45 Binaural interactions in short-term adaptation were investigated in IE units by examining the magnitude and time-course of changes in discharge rate during 200 ms binaural adapting tones over a range of stimulus intensities. Ipsilateral and contralateral intensities were manipulated to delineate possible relationships between IIDs and short-term adaptation. Responses o f a typical binaural L S O unit to binaural adapting stimuli without interaural intensity disparity (IID = 0) are depicted in P S T H s of Figure 4.05 A . Increasing the overall intensity of the binaural stimulus (IID = 0) resulted in slightly greater discharge rates over the 200 ms (adapting) stimulus (Figure 4.05 A , bottom PSTH to top), but without a marked change in the time course o f the response. That is, the P S T H s were flat, and very little short-term adaptation was evident. This was observed for all levels of the binaural stimulus where interaural intensities were equal. Short-term adaptation has, therefore, no major influence on binaural L S O neuron responses provided interaural stimulus intensities are matched (IID = 0). During binaural stimulation where interaural intensities were not equal (IID * 0), an emphasis on response onset characterized the P S T H . Small IIDs resulted in the emergence o f an small early peak in the P S T H (Figure 4.05 B , middle PSTH). A s the ipsilateral stimulus level was raised relative to the contralateral stimulus, steady-state discharge rate increased, and the P S T H onset peak became more prominent (cf. Figure 4.05 B , PSTHs). Short-term adaptation was therefore evident as response decrements over the stimulus period. For the typical unit depicted in Figure 4.05, the magnitude of short-term adaptation grew from near zero levels to 81 %. In addition, the time constant o f short-term adaptation was shortened from 6.5 ms (IID = 19 dB SPL) to 3.2 ms (IID = 28 dB SPL). Therefore, IID has a complex effect on spike discharge in binaural L S O units. While steady-state (plateau) responses did increase with IID, peak responses showed much T. J . A D A M 46 greater sensitivity, becoming more prominent as the IID was raised. Evidently, short-term adaptation grew with the IID. Units consistently showed this response profile during manipulations o f binaural stimuli. During equal-intensity binaural stimulation (IID = 0), the average spike discharge rate was stable throughout the binaural stimulus (n=12), but increased with greater overall stimulus intensities. During unequal-intensity binaural stimulation (IID * 0), early onset peaks appeared in P S T H s and short-term adaptation was evident (n=19). For all IIDs tested, the magnitude o f adaptation was, on average, 62 ± 7.1 %, and occurred with an average time constant of 6.9 ± 2.3 ms. However, as with manipulations of ipsilateral intensity during monaural stimulation, these parameters varied with stimulus level difference (IID, A N O V A p<0.05). In the IID range o f 0 to 28 dB, the magnitude of adaptation increased from 0 to 71 ± 18 %, with a time constant approaching 4 ± 0.9 ms (n=19). These results demonstrate that short-term adaptation occurs in responses o f L S O IE neurons only i f evoked by binaural stimulation with interaural intensity disparities (IIDs). Equal interaural intensities (IID = 0) result in relatively stable but low firing rates throughout the stimulus period. For unequal interaural intensities (IID * 0), spike discharge is "never stable throughout the stimulus period. Firing rate is initially high ( P S T H peak), and gradually decreases over time to. a steady-state level ( P S T H plateau). O n average, binaural responses are decreased by 62 %, with a time constant o f 7 ms. Furthermore, the size o f the IID affects short-term adaptation, determining both its magnitude and time course. Greater IIDs result in greater short-term adaptation along a faster time course. Therefore, contrary to prevailing assumptions, no un-ambiguous code o f IID or azimuthal sound incidence angle exists in the steady-state firing rate o f L S O IE neurons, in the amplitude o f its onset, or in the entire responses magnitude. Individual T. J. A D A M 47 L S O cells cannot distinguish, by firing rate alone, between sound source position, movement, and intensity changes. 6.2.4 Post-stimulatory effects of short-term adaptation in IE units Short-term adaptation in excitation and inhibition of L S O neurons has been demonstrated, above, during prolonged stimulation. The effects of short-term adaptation are expected to influence responses to subsequent stimuli. We hypothesized that short-term adaptation invoked by prolonged stimuli w i l l influence responses to subsequent tones, and predicted that post-stimulatory effects wi l l wane with time from the adapting stimulus, until the responses recover to their initial levels. Again, the hypothesized rate code o f IID requires that recovery for excitatory and inhibitory short-term adaptation is balanced. In the following experiment, we examine the post-stimulatory effects of short-term adaptation on L S O EE neuron discharge rate and onset latency. The extent and time course o f recovery was examined by comparing the discharge rate during 30 ms probe tones presented at various delays from the termination o f a 200 ms adapting tone. For all units studied, the amplitude of responses to probe tones at a 512 ms delay was not statistically different from the amplitude o f responses to the initial 30 ms of the adapting stimulus. We normalized the adapted responses, therefore, to these control responses Recovery from short-term adaptation was examined for ipsilateral, contralateral, and binaural stimulation by comparing probe responses to those obtained at the 512 ms delay. 6.2.4.1 Equal IEDs Recovery from short-term adaptation is illustrated in Figure 4.06 for a typical binaural L S O unit. Response magnitude during short-term adaptation and recovery in ipsilateral excitation, contralateral inhibition, and binaural stimulation are presented in T. J. A D A M 48 Figure 4.06 A - C , respectively. Responses (PSTHs) during the adapting tones are shown in the left column, responses to probe tones presented 1 ms following the adapting tone are shown in the middle column, and recovered responses (interval = 256 ms) in the right column. For monaural stimulation, persistence o f short-term adaptation following the adapting tone is evident in PSTHs , and the time course of recovery is evident in responses to probe tones at longer delays. In A , it is clear that excitatory short-term adaptation occurred during the ipsilateral adapting tone (left), and persisted, as decreased spike discharge, during the first ipsilateral probe tone (middle). Ipsilateral responses recovered within 256 ms, as illustrated by probe responses that match the amplitude (right) to the initial 30 ms of the adapting tone. Inhibitory short-term adaptation imposed through the presentation of a contralateral adapting tone also outlived the stimulus, as indicated by relatively greater spike discharge during a binaural probe tone presented 1 ms after the adapting tone (Figure 4.06 C). However, binaural responses (IID = 0) remained stable after the adapting tone (Figure 4.06 B) . In order to assess the time course of recovery from short-term adaptation, the ensemble discharge rate the 30 ms probe was plotted against probe delay from the terminus of the adapting tone. This is illustrated in Figure 4.06 D for the same typical unit. The time constants of recovery from short-term adaptation in excitation (62.7 ms, 17 dB SPL, fdled circles) and inhibition (51.1 ms, 17 dB SPL, fdled triangles) were similar, resulting in mirror-imaged recovery functions. Binaural response amplitudes (IID = 0) remained low and stable after the adapting stimulus (unfilled squares). This was the typical recovery profile for all neurons that could be tested with equal interaural stimulus levels (n=12). In each case, adapted excitation and inhibition recovered along similar time courses, and binaural responses were stable throughout. Consequently, stability o f binaural responses may depend on balanced short-term adaptation and recovery from excitation and inhibition converging in the L S O neuron. T. J. A D A M 49 6.2.4.2 Unequal IIDs In L S O neurons, only positive IIDs (ipsilateral tone louder) could be tested with an ipsilateral stimulus of sufficient intensity to produce a robust spike discharge. The contralateral stimulus intensity was then varied to produce a range o f IIDs. Therefore, the ipsilateral stimulus was more intense than the contralateral intensity in each trial. A s a consequence of the relationship between short-term adaptation and stimulus intensity (above), greater post-stimulatory short-term adaptation magnitudes were expected for excitation than inhibition. This was evident in the recovery functions of monaural excitation and inhibition (Figure 4.07). Excitatory short-term adaptation produced greater response changes following the stronger ipsilateral adapting stimuli than inhibition did following the weaker contralateral adapting stimuli (e.g. Figure 4.07 A ) . Further, the degree o f asymmetry in monaural recovery functions was directly related to the relative stimulus intensities used. Comparing monaural recovery functions in Figure 4.07 A , versus Figure 4.07 B , reveals that similar ipsilateral intensities (40 versus 42 dB SPL) led to similar post-excitatory short-term adaptation magnitudes. Conversely, reduced inhibitory stimulus intensities (35 versus 17 dB SPL) led to reduced post-inhibitory short-term adaptation. Therefore, in parallel with the induction of short-term adaptation (excitation Figure 4.03, inhibition Figure 4.04), post-stimulatory recovery functions were directly related to respective stimulus intensities, and therefore IID ( A N O V A , p<0.05, n=19). The time course o f recovery from short-term adaptation during monaural stimulation was, however, similar for excitation and inhibition, and did not vary significantly with stimulus intensities. The recovery from monaural excitatory adaptation tended to be faster, although inhibitory recovery was much more variable (excitatory time constant: 53.5 ± 1 1 ms, inhibitory time constant: 107 ± 81 ms, A N O V A p<0.05, n=T9). Therefore, the absence of symmetry in recovery functions of excitation and inhibition T. J. A D A M 50 obtained in each case (Figure 4.07 A , B ) was chiefly due to greater post-stimulatory effects of the more intense excitation than inhibition, and this influenced the magnitudes of post-stimulatory adaptation effects. Together, this evidence suggests that the magnitude o f EE neuron discharge following binaural short-term adaptation wi l l recover in relation to the intensities of excitation and inhibition (IID). Due to the relatively greater stimulus intensity for excitation during positive IIDs, reduced binaural discharge should be apparent following binaural stimulation. Binaural responses should then recover to higher control levels along a time course similar to those following monaural excitation. Furthermore, post-stimulatory effects o f binaural short-term adaptation should be related to the magnitude of IID. This would be consistent with our presumption that the summation o f excitation and inhibition determines the IE neuron discharge rate during binaural stimulation. For each unit tested with a positive IID, decreased responsiveness was evident following binaural short-term adaptation (e.g. Figure 4.07 A ) . Increased discharge rates at longer probe delays reflected recovery o f binaural responses. Therefore, the post-stimulatory effects o f binaural short-term adaptation were related to greater short-term adaptation in excitation than inhibition, as predicted. However, there was no systematic relationship between the magnitude o f this effect and the IID. Comparison o f binaural recovery in Figure 4.07 A and Figure 4.07 B reveals that lower post-stimulatory recovery effects after binaural short-term adaptation could be obtained with larger IIDs. Over the sample of binaural units tested, there was no significant relation between IID and the amount o f post-stimulatory recovery from binaural short-term adaptation (n=19). A n unexpected finding was that the time course o f recovery for binaural discharge rates following binaural short-term adaptation was not related to the time courses of recovery from either monaural excitation or inhibition. Following stimuli with moderate to large IIDs, binaural recovery appears to be slower than expected in Figure 4.07, particularly B . Over our sample, the recovery o f binaural responses did not reflect the T. J. A D A M 51 recovery from either monaural excitation or monaural inhibition (n=19). Thus, based on excitation and inhibition parameters, only the qualitative prediction that binaural responses were decreased following binaural short-term adaptation was realized. The magnitude of these response decrements, and their time course, were not related to excitatory or inhibitory recovery functions. Therefore, the post-stimulatory recovery from binaural short-term adaptation also seems not to be systematically related to IID. In summary, post-stimulatory effects of short-term adaptation are apparent in excitation and inhibition. During balanced binaural stimulation (IID = 0), responses are stable. Conversely, short-term adaptation resulting from binaural stimulation with disparate interaural intensities (IID * 0) results in decremented binaural discharge rates, which recover as the delay between the adapting and test stimuli increase. However, post-stimulatory recovery from binaural stimulation was not related to IID, even though recoveries from monaural excitation and inhibition were. Discharge rates decrease following binaural short-term adaptation with positive IIDs. This post-stimulatory rate decrease, and its recovery, are not related to IID in any systematic manner. Therefore, a summative integration o f adapted excitation and inhibition does not predict the observed changes in responsiveness following binaural stimulation. The hypothetical rate code for IID in individual IE neurons would appear to be seriously degraded during IIDs resulting from most normal sound source directions. 6.2.5 The influence of short-term adaptation on response onset latency In previous sections, we noted that the onset latency of IE neuron discharge was related to the intensity o f the ipsilateral stimulus. Unless contralateral stimulus intensities suppressed spike discharge altogether, onset latencies were unaffected by inhibition. We now return to examine the possible influence o f short-term adaptation on the onset latencies. The latency o f probe responses is measured as a function of probe delay to T. J. A D A M 52 determine whether excitatory short-term adaptation influences the latency of firing in subsequent responses. In addition, the possible influence of binaural short-term adaptation on the latency is examined. Results for a typical EE unit are depicted in Figure 4.08 under conditions of equal and unequal interaural intensities ( A ands B , respectively). Large changes in probe response latency were evident following excitatory short-term adaptation. Immediately following the adapting stimulus, latencies were approximately doubled, and returned to pre-adaptation levels over the same time course that discharge rate recovered (cf. Figure 4.06 D) . Responses latencies during probe tones following inhibitory short-term adaptation remained at control levels (3.9 ms - response latency during an ipsilateral adapting tone o f equivalent intensity). Over the sample of binaural units examined (n=19), onset response latency was significantly related only to the recovery of excitation ( A N O V A , p<0.05). Binaural short-term adaptation also impacted response onset timing, in a manner similar to that of excitation. During monaural and binaural stimulation, inhibition appeared to bear little influence on the onset latency of EE neuronal discharge. Therefore, response onset latency is apparently determined only by ipsilateral stimulus intensity, and the adaptation state of the excitatory input to L S O EE units. This result is consistent with our observation that inhibition converges on the L S O neuron to preferentially influence later portions of the chopper response. Clearly, the integration of excitation and inhibition in EE L S O neurons is more complex than previously expected. While chopper discharge rate is related to the relative intensities at the ipsilateral and contralateral ears, the onset latency is determined only by the former o f these. This suggests that the influence of inhibition on chopper discharge may show temporal dependence, to affect the discharge rate of chopper responses only. Short-term adaptation in monaural units of the superior olivary complex T. J. A D A M 53 The excitatory inputs of binaural IE units originate in spherical bushy cells of the ventral cochlear nucleus ( V C N ) . These neurons are monaural, responding to ipsilateral stimulation with excitation (OE). The inhibitory input to the L S O is comprised of principal neurons of the M N T B which receive excitatory input from the V C N globular bushy cells. The inhibitory input neurons respond to contralateral stimulation (EO). Short-term adaptation has been demonstrated in the V C N neurons (Boettcher, Salvi and Saunders, 1990; Shore, 1995; Watanabe and Simada, 1971). We sought to determine the magnitudes and time constants of short-term adaptation in those monaural units that likely constitute the L S O inputs, in order to relate them to the adaptation in IE neurons. I f short-term adaptation observed in L S O neurons is simply imported from their inputs, its time course and magnitude should match those of monaural input neurons. Short-term adaptation of monaurally driven units in the S O C was measured and compared to short-term adaptation in binaural units. Significantly different magnitudes and time courses of short-term adaptation would support contributions from input synapses and/or intrinsic properties o f the IE L S O unit. Units of the SOC were considered to be monaural i f they were responsive to stimulation of one ear, and concurrent stimulation of the other ear had no effect on this excitation (OE, EO). O f all units studied in the SOC, 44 satisfied this criterion. However, of this sample, only 38 units exhibited primary-like discharge patterns (e.g. Figure 4.09 A , left P S T H ) . Only this population is o f interest in the present consideration, since they likely comprise the primary input neurons to the L S O IE neurons. Consequently, only monaural units exhibiting primary-like discharge patterns are considered here (n=38). For monaural units in the S O C , short-term adaptation was examined using the same stimulation and analysis protocols employed for the examination o f excitatory short-term adaptation in binaural L S O units. The magnitude of decrements in spike discharge, and the time constants of exponential functions fit to P S T H s , and to the magnitudes of probe responses during recovery, characterized short-term adaptation in monaural units. For these units, the parameters of short-term adaptation did not vary T. J. A D A M 54 systematically with aurality (OE or EO), or with the location within the SOC. Consequently, data were pooled for all monaural units exhibiting primary-like responses. The general characteristics of short-term adaptation are illustrated for a typical monaural unit in Figure 4.09 A (an OE cell, B F = 25.05 kHz) . During a 200 ms adapting tone, a P S T H shows a marked decline in the response as the stimulus proceeds (Figure 4.09 A , left). Spike discharge rate decreased from an early pronounced peak in the P S T H to a steady-state level. This is indicative of short-term adaptation in excitation. In this example, the magnitude o f short-term adaptation was 64%, and occurred along a time course with the exponential time constant of 5.14 ms (the exponential function fit to the P S T H is shown offset). Peri-stimulus time histograms for responses to probe tones following the adapting tone show an early decline in magnitude and recovery as the delay from the adapting tone increases (Figure 4.09 A , middle and right). B y a delay of 256 ms, the pre-adaptation response magnitude (average spikes/bin over 30 ms response) was achieved, although the onset peak is spread over a longer duration. A n exponential fit to recovery data at this stimulus intensity (40 dB SPL) possessed a time constant o f 15.7 ms. A s expected, all 38 monaural units displayed short-term adaptation during tonal stimulation, and recovery therefrom. Results depicted in Figure 4.09, and the indices of short-term adaptation obtained for that unit were typical for monaural units examined. The short-term adaptation resulted in overall response decrements o f 48.2 ± 4.9 %, with time constants o f 9.2 ± 1 . 7 ms. Recovery from short-term adaptation was nearly complete within 100 ms, as indicated by the average recovery time constant of 72.2 ± 23.6 ms. This was significantly slower than the time course of short-term adaptation (p<0.01). Surprisingly, the magnitude and time course of short-term adaptation were not significantly related to stimulus intensity in monaural units. Although the unit depicted in Figure 4.09 showed shortening o f the recovery time constant from 43.0 ms to 15.7 ms over a 10 dB SPL intensity change, this trend did not generalize to the sample o f units T . J . A D A M 55 examined (p>0.05). Furthermore, the magnitude and exponential time constant of short-term adaptation during the 200 ms stimulus could not be related to intensity (p>0.05). A l l of these indices remained relatively stable with changes in stimulus level from 12 dB SPL to 40 dB SPL. Consequently, short-term adaptation and recovery in monaural units does not appear to vary significantly with this intensity range. Relative to binaural L S O units, the magnitude and time course o f short-term adaptation was different in monaural SOC neurons. Response decrements to ipsilateral excitation were significantly greater (68 % versus 48 %, p<0.05), and occurred more quickly (5.11 ms versus 9.24 ms, p<0.05) in binaural units. However, the time course of recovery did not deviate significantly from monaural units. This is likely due to the relatively large variances o f recovery time constants observed in both types of unit, which resulted in decreased statistical power. Our results, therefore, support a contribution by synaptic and/or intrinsic properties o f L S O IE neurons to short-term adaptation, in addition to the response decrements already present in the primary inputs of the L S O . T. J. A D A M 56 7. E X A M I N A T I O N OF INTRINSIC M E M B R A N E PROPERTIES IN NEURONS OF THE LSO 7.1 Introduction In the previous section, it was shown that the discharge pattern o f IE L S O neurons is shaped to large extent by their own properties. We now turn to an experimental investigation o f intrinsic membrane properties contributing to the adapting chopper responses in IE L S O neurons. We describe voltage measurements during direct current injection o f L S O chopper neurons maintained in the tissue slice preparation, in vitro. Voltage responses characterize the chopper neuron with respect to its firing pattern and rectifying properties subthreshold to action potential generation. This evidence also suggests that the integration of excitation and inhibition is more complex than expected. It has been suggested previously that the preferential influence of inhibition on later portions of the chopper response is due to the delayed arrival o f this input at the L S O neuron, relative to excitation (Tsuchitani, 1988, a, b). A s the contralateral stimulus intensity is raised, the resultant synaptic volley arrives earlier to desynchronize and truncate earlier portions of spike discharge. However, intracellular recordings of IE units in the L S O indicate that inhibition actually arrives earlier that excitation at the L S O spike generation zone (Finlayson and Caspary, 1993). Taken into consideration with our present findings, inhibition may in fact arrive before excitation to set the L S O cellular membrane potential from which excitatory synaptic volleys determine the chopper response. This possibility immediately calls into question the potential role for intrinsic membrane properties of the L S O chopper neuron in (i) the integration o f excitation and inhibition, and (ii) the generation o f chopper responses. Were inhibition to hyperpolarize the chopper membrane, and increase membrane conductance prior to the arrival of the T. J. A D A M 57 excitatory synaptic volley, intrinsic membrane properties may serve to compensate for inhibitory effects o f inhibition on chopper responses, and short-term adaptation effects during binaural stimulation. This hypothesis led us to the following series o f experiments. The primary aim was to delineate the repertoire of voltage-dependent membrane conductances contributing to chopper response generation, and its possible modulation by membrane potential. Intrinsic membrane conductances must mediate the preservation o f chopper onset in the face o f variant stimulation history (i.e. inhibitory synaptic volleys). The intrinsic membrane properties of chopper neurons were delineated in the tissue slice, and then their voltage dependence determined through the manipulation of holding potential (membrane voltage at the onset o f the stimulus). I f our predicted integration scheme is correct, chopper neuron behaviour described in vivo should be mimicked in the tissue slice through the hyperpolarization o f holding potential (inhibition) and subsequent depolarizing current pulses (excitation). Before proceeding, a brief overview o f the experiments reviewed in this chapter is in order, (i) The cellular morphology o f cell types encountered in the L S O were related to basic membrane properties and spike discharge (n=13). This experiment identified chopper-like responses as belonging to L S O principal neurons, and delayed firing to descending efferents residing in the L S O . (ii) The spike discharge behaviour was described for L S O principal neurons in the tissue slice (n=110). These experiments related the discharge behaviour of L S O neurons to findings reported in vivo, (iii) Intrinsic membrane properties contributing to chopper-like discharge in the tissue slice were delineated, and contributing membrane conductances identified (n=126). (iv) The influence o f membrane potential on the chopper-like responses was examined using current prepulses (n=25). (v) The spatial and temporal summation o f synaptic potentials in L S O principal neurons were investigated (n=17). (vi) The temporal response properties and spike discharge statistics for the efferent neurons o f the lateral olivocochlear system were examined (n=34). T. J. ADAM 58 7.2 Methods The primary purpose of this series of experiments was to characterize the intrinsic mechanisms of chopper response generation and short-term adaptation in L S O principal neurons. A tissue slice preparation of the rat auditory brainstem was developed to this end. Intracellular recordings were conducted to characterize the chopper response as it occurs in vitro, and describe relevant intrinsic membrane properties. Whole cell patch recordings were performed to identify membrane conductances underlying these intrinsic properties. 7.2. / Selection of animals Juvenile Long Evans rats, aged 11 to 14 days, comprised the subject base for tissue slice experiments. Juvenile animals were chosen for their relatively soft skulls, and consequent ease and rapidity o f brain extraction. Al so , younger brain tissue is more resistant to the effects o f hypoxia during surgical preparation. Finally, adequate seals between the electrode and neuronal membrane needed for patch clamping are difficult or impossible to obtain in mature brainstem slices. This is likely due to the onset of myelination in brainstem fibres around day 14. Recordings conducted in animals under 11 days of age revealed significant age-dependent membrane behaviour, and were therefore not included in this dissertation. O f course, considerable development in auditory functioning occurs after 14 days but, by day 12, auditory thalamic neurons o f the rat are mature (Tennigkeit, Schwarz and Pui l , 1998). Auditory neurons in the lower brainstem nuclei, such as the L S O , probably mature slightly earlier. Additionally, behavioural auditory thresholds (measured via the orienting reflex) are within the adult range, and our informal behavioural observations indicate T. J. A D A M 59 directional hearing capabilities in all fats used in these experiments (Adam, 1992). It is reasonable, therefore, that the membrane properties of L S O neurons support normal auditory function at this age range. 7.2.2 Surgical preparation of rat auditory brainstem slices The tissue slice preparation has proven to be a highly successful tool for the study of intrinsic membrane, morphological, and synaptic properties of neurons in the central nervous system. Tissue slices o f the rat auditory brainstem were employed to study the cellular properties o f L S O neurons for a number of advantages it offers over the anaesthetized animal. M o s t importantly, the tissue slice eliminates influences from other regions of the brain. Responses therefore reflect the intrinsic electrophysiology o f the target structure, intra-slice circuitry, and stimulus conditions. The tissue slice also offers experimental control over the external milieu: Extracellular salts and pharmacological agents are easily manipulated and allowed direct access to the tissue (Alger, 1984). Finally, stable intracellular and whole cell patch recordings are easier to obtain in vitro than in the anesthetized preparation due to the elimination o f brain movements associated with breathing, for example. The protocol developed for slice preparation o f the auditory brainstem enhanced the viability of the tissue by: (i) reducing metabolic demands of constituent neurons, (ii) reducing neuronal activity, and (iii) entailing a rapid and careful dissection to minimize mechanical damage. Such emphasis is particularly important for slice preparation of brainstem regions where metabolic demands are high. The brain was rapidly cooled to 4-6 °C before dissection, and sucrose was substituted for sodium in the external (tissue bath) fluid. This reduced spike discharge during tissue manipulations, thereby decreasing the probability o f triggering the cascade of intracellular reactions associated with excitotoxicity. Tissue viability was greatly enhanced by these procedural measures. T. J. A D A M 60 A t the outset of the experiment, rats were rapidly anaesthetized with the inhalatory agent, halothane, and decapitated. The scalp was then bisected and reflected laterally. The occipital condyles were cut, and the cranium bisected longitudinally along the midline, from the foramen magnum. The skull was cut laterally at the level of frontoparietal sulcus, which was visible through the skull. The two halves of the skull cap were deflected laterally to expose the brain, and the dissected forebrain removed. The preparation was then submerged in ice-cold sucrose-containing artificial cerebrospinal fluid saturated with carbogen (95% 0 2 / 5 % C 0 2 ) . The brain block was extracted, with careful transection of the cranial nerves to prevent mechanical damage to fibre tracts. This dissection took approximately 50 seconds. A transverse tissue block o f the brainstem, approximately 6 mm thick, was dissected with a razor blade; removing first the entire forebrain, then the cerebellum (including flocculi), and remaining spinal cord. This block contained the trapezoid body, which was easily visible as a white band on the ventral surface of the brainstem. The caudal surface of the block was blotted to remove excess fluid and glued (cyanoacrylate glue) in place on a vibratome (Pelco 101 Series 1000) stage for slicing. The block was submerged in ice-cold standard artificial cerebrospinal fluid ( sACSF) , and coronal slices (300-350 ( i M thickness, similar to the schematic of Figure 3.01) were taken. Those containing the L S O were transferred to a holding chamber filled with A C S F , maintained at 30 °C, and bubbled with carbogen. After slicing, the fluid and slices in the recording chamber were cooled to room temperature (22 °C), and stored for one hour, prior to the recording session. For experiments conducted at 34 °C, the fluid in the recording chamber was warmed to this temperature, and the slices were acclimated briefly following transfer to this setup. Slices were generally viable at 22 °C for 4-6 hours, and 1-3 hours at 34 °C. Procedures and protocols conducted on animals were approved by the Canadian Council on Animal Care ( U B C protocol number: A95-0275). T. J. A D A M 61 7.2.3 The recording chamber and associated equipment The recording chamber (Figure 5.01) optimized a number of features critical for the viability of the tissue slice and recording conditions. Tissue slices were suspended on nylon mesh fixed approximately 1 mm above the bottom of the chamber. They were held in place with a "slice harp" comprised o f nylon thread stretched and affixed across a U -shaped piece o f platinum wire. This arrangement placed the tissue slice in the midst of A C S F flow (Figure 5.01), maximizing permeation of the slice by external solutions and oxygen. A brisk flow rate (4-6 ml/minute) supported the high metabolic demands of auditory brainstem tissue, and ensured rapid fluid exchange and equilibration of external solutions. A t the origin of the fluid delivery system, the A C S F was continuously bubbled with carbogen. A roller pump (Cole Parmer, Masterflex C / L ) propelled the A C S F to the recording chamber at a constant rate of 4-6 ml/minute, minimizing changes in fluid level during A C S F exchange (Haas, et al., 1992). To further eliminate pressure pulses in A C S F flow, a syringe, partly filled with A C S F and air, was connected to the fluid line, creating a T-type connection. After bathing the slice, fluid exited the recording chamber, and was removed via suction from a connected reservoir (Figure 5.01). Through height adjustments of the suction pipette, the recording bath fluid level could be controlled. The level was set so that tissue slices were barely submerged in the meniscus of A C S F . This reduced the electrode capacitance associated with bath fluid depth, while minimizing bath volume (0.7 ml), which permitted rapid drug equilibration and washout. For experiments conducted at 34 °C, oxygenated A C S F was warmed to 36-37 °C in a warm water bath (Precision Scientific Inc. temperature regulator, Mode l #182, Chicago, IL.). The slightly warmer temperature of the water bath ensured a lower saturation level o f carbogen, preventing gas bubble formation in the recording chamber T. J. A D A M 62 (and under the slices). A thermoregulator ( A T R - 4 , Quest Scientific) and placement of the thermistor in the recording chamber ensured appropriate temperature maintenance. The recording chamber and electrode holder were mechanically and electrically isolated from their surroundings with a vibration isolation table (Newport) and a Faraday cage. 7.2.4 Solutions and electrodes Depending on the general purpose of the experiment, the external (bath) and internal (electrode) solutions were manipulated to achieve various goals. In cases where in vitro chopper neuron responses were examined, sharp microelectrodes were used, in order to minimize diffusion into the cellular interior. Other experiments were conducted with the aim o f placing the cell interior under experimental control. In these instances, patch electrodes with relatively large tip openings were employed. 7.2.4.1 External solution The standard A C S F contained (in m M ) : 125.0 N a C l , 2.5 m M K C I , 2.0 C a C l 2 , 2.0 M g C l 2 , 1.25 N a H 2 P 0 4 - H 2 0 , 10.0 6-glucose, and 25.0 N a H C 0 3 . When saturated with 95% 0 2 / 5 % C 0 2 , the p H of the A C S F was 7.3. This composition o f yielded a measured osmolarity of 312 mOsmoles. 7.2.4.2 Internal solution for sharp microelectrodes Sharp microelectrodes were pulled from borosilicate glass (1.2 mm outer diameter, World Precision Instruments, Sarasota, F L ) with a horizontal puller (Sutter Instruments, Mode l P-87). The intracellular solution was either K C I (3 M , 60-70 M Q ) or K-acetate (2 T. J. A D A M 63 M , 80-100 M Q ) , titrated to p H 7.3-7.4 using HC1 or K O H for the former, and acetic acid o r K O H for the latter. 7.2.4.3 Internal solution for patch electrodes Patch electrodes were pulled from borosilicate glass (1.5 mm outer diameter, World Precision Instruments Inc., Sarasota, F L A ) using a two-stage vertical puller (Narashige Scientific Instruments, Mode l PP-83). The electrode filling solution for whole cell patch recordings was composed o f (in m M ) : 115 K-gluconate, 20 K C I , 1 C a C l 2 , 10 ethylene glycol-bis(r>aminoethylether) N,N,N',N',-tetraacetic acid ( E G T A ) , 10 Na-N-2-hydroxyethylpiperazine-N'-2-ethanesulfonate (HEPES) , 4 magnesium adenosine triphosphate ( M g - A T P ) , 0.3 sodium guanosine triphosphate (Na-GTP) . Free internal calcium concentration associated with this electrode solution was estimated to be 1 x 10"8 M (MaxChelator 1.2, Pacific Grove, C A ) . This solution was titrated to p H 7.3-7.4 using 5-gluconic acid or 2-amino-2-methyl-l-propanol. Electrode resistances were 4-6 M Q . For whole-cell recordings, where relatively free access is achieved between the internal and external solutions, the equilibrium potentials were calculated using the Nemst equation (Hille, 1992; Johnston, 1995). The formula and values for each ion are presented in Table 5.01. 7.2.5 Intracellular recordings Using a Newport motorized actuator (Model 850B) and controller (Model # P M C 100), electrodes were advanced into the L S O , which was identified visually under a dissection microscope. Cells were typically encountered between 50 and 150 um from the tissue surface. For each recording, voltage responses to direct current injection were recorded with an Axoclamp 2B amplifier (Axon Instruments, Sarasota F L ) in current clamp. Data T. J. A D A M 64 were low-pass filtered at 3 k H z , amplified (Kikisui Instruments, Mode l COS5020) and sampled at 20 k H z (National Instruments board: M I O - 1 6 H ; D M A - 2 8 0 0 ) in a Macintosh Quadra 650 computer controlled by A/Dvance software (McKel lar Designs, Vancouver B C ) . Prior to each experiment, the access resistance and electrode capacitance were compensated for using circuitry supplied in the Axoclamp amplifier, while the electrode was in the bath. Access resistance was compensated through the utilization o f internal wheatstone bridge. Electrode capacitance was compensated for using capacitance neutralization circuitry. A voltage offset was applied to the recording to bring bringing the amplifier to a zero level when the electrode tip was in the bath. However, no attempt was made to correct for further junction potentials between the electrode solution and cell's microenvironment (Muller, 1992; Sherman-Gold, 1993). This limited the precision o f our membrane potential measurements. 7.2.5.1 Sharp microelectrodes Sharp microelectrodes were employed to characterize the intrinsic membrane properties, spike discharge, and corresponding statistics of L S O neurons. A s well , they were used to examined short-term adaptation effects on postsynaptic potentials in the L S O principal neuron. Repetitive search stimuli (0.3 n A , 40 ms. duration) were presented (intertrial interval 0.3 seconds) while advancing the electrode through the slice. A n increase in resistance produced a voltage deflection of ~ 3 m V indicated that the electrode had advanced into the vicinity of a presumptive cell. A t this point, breakthrough into the cell was achieved by a brief overutilization of the capacitance neutralization circuit (Muller, 1992). Negative direct current (up to 0.20 n A ) was then applied to the cell, bringing the holding potential below -60 m V , until the recording stabilized (firm holding potential, no spontaneous firing). Steady current was then removed, and the recording session began. T. J. A D A M 65 7.2.5.2 Patch electrodes Patch recordings in the whole cell configuration offered a number of advantages over the intracellular configuration: The greater current passing capabilities of patch electrodes allowed for easier manipulation of holding potentials, and voltage dependence of L S O neuron behaviour was easier to study in current clamp. Electrode advancement was performed as with microelectrodes. However, slight positive pressure was applied to the electrode solution, keeping the electrode tip free of debris. A blind approach was taken through the L S O , until initial contact with a presumptive cell was indicated by a slight increase in the voltage deflection during search stimuli (0.43 n A , 40 ms duration). Negative pressure was then applied, and the search stimulus was gradually dropped to 0.03 n A , as electrode resistance increased until > 1 G O (typically 3 G Q ) seal was obtained between the electrode tip and the cell membrane. A sharp negative pressure pulse resulted in breakthrough to the cell interior and establishment of the whole-cell recording configuration. Recordings commenced 5 minutes later, after the internal electrode solution was, presumably, at an equilibrium with the interior of the cell. 7.2.6 Current clamp recordings Data reported were obtained from L S O neurons maintained at stable membrane potentials for recording sessions that exceeded one hour. A l l voltage recordings were conducted in current clamp: Current pulses (200 ms duration) of known amplitude were injected into the cell directly via the recording electrode. Respondent voltage traces were simultaneously measured using the same electrode. Electrophysiological responses were initially recorded with no constant current application. During patch recordings, the holding potential was subsequently manipulated through the introduction of steady direct T. J. A D A M 66 current through the recording electrode. Current pulses were then superimposed. A t least three holding potentials were tested: one in the absence o f steady direct current ("rest"), one at a relatively depolarized baseline level (-10 m V positive to rest), and one at a hyperpolarized level (-20 m V from rest). Inter-stimulus intervals and constant current were manipulated manually to ensure a stable baseline membrane potential. This protocol allowed us to describe certain intrinsic membrane properties of the neuron from the voltage responses to current injection. 7.2.6.1 Basic membrane properties Basic membrane properties were measured directly from voltage responses evoked from the resting membrane potential ( R M P ) . These included R M P itself, action potential amplitude, spike half-width, and threshold voltage. The membrane time constant was derived from single exponential fits to small amplitude hyperpolarizing voltage responses. Input resistance was determined from onset voltage responses during hyperpolarizations of less than 10 mV. Current-voltage (I-V) relations were measured for the peak (largest magnitude voltage deflection in a response) and steady state (195 ms) segments o f voltage responses. These basic membrane properties were compiled from recordings conducted with sharp and patch electrodes. 7.2.6.2 Spike discharge patterns In order to characterize the firing pattern and discharge statistics (e.g. firing rate, discharge regularity), repeated stimulus trials were required for analysis in each cell. Depolarizations o f a given amplitude and 200 ms duration were presented repeatedly (50 trials) with an inter-trial interval o f 3 to 5 seconds. Voltage responses were recorded with sharp microelectrodes during stimulus presentations. To assess the relation o f discharge statistics to current magnitude, three or more depolarization levels were studied. This T. J. A D A M 67 protocol permitted us to study the relation between discharge statistics and stimulus amplitude and onset timing. 7.2.6.3 Synaptic potentials A series o f experiments was conducted in order to address synaptic depression in L S O principal neurons. Evokation of synaptic potentials was achieved with a single bipolar stimulating electrode (0.1 mm tip diameter, World Precision Instruments) driven by two stimulation isolation units (Neurolog) and controlled by A/Dvance (McKel lar Designs). Excitatory postsynaptic potentials (EPSPs) were reliably elicited when the stimulating electrode was placed on the surface of the tissue slice, 1-2 mm lateral to the L S O to stimulate the ventral acoustic stria. The distance from the L S O was 1 to 5 mm. Inhibitory postsynaptic potentials (IPSPs) were evoked when the stimulating electrode was placed medial to the L S O , in the M N T B . Single, bipolar, small-amplitude current pulses (50-200 u A , 100 us) applied to the stimulating electrode reliably evoked EPSPs and IPSPs. Initially, single pulses of varying amplitude were delivered to the L S O input tract in order to establish a moderate stimulation level for an assessment o f synaptic depression in postsynaptic potentials. Short-term synaptic depression was addressed in these cells by presenting a short (10 ms) stimulus train of 10 pulses (adapting stimulus) followed by a probe pulse at a variant delay (1 ,2 ,4 , 8,16, 32, 64, 128,256 ms). Stimulus intensities were set to produce moderately-sized PSPs. Synaptic depression ofPSPs in chopper neurons of the L S O could then be compared to the short-term adaptation observed in vivo. 7.2.7 Ion channel antagonism Ion channel blockers were applied to the external milieu to investigate the contribution o f particular ion channel families to subthreshold and suprathreshold membrane behaviour. Effective dosages were adopted from the literature or determined in T. J. A D A M 68 cumulative dose-response studies. Reversal o f pharmacological (wash out) was obtained whenever possible, and new slices were employed for each pharmacological manipulation, unless otherwise stated. Where pharmacological agents shifted the resting potential, constant current was applied to restore the holding potential so that comparisons could be made to control levels. Input resistances were derived from linear sections of the I -V relation under different pharmacological conditions. Changes in measurements with experimental condition are given as percentages ± standard deviations. Ion channel blockers employed included: 300 to 600 n M tetrodotoxin ( T T X ) , 50 u M C d 2 + , 50 u M N i 2 + , 0.05 to 4 m M 4-aminopyridine (4-AP), 0.5 to 20 m M tetra-ethylammonium ( T E A ) , 0.2 m M B a 2 + , and 3 m M C s 2 + . In one series o f experiments, external calcium concentration was reduced to 200 u M , with overall divalent cation concentration being maintained with M g 2 + . 7.2.8 Spike discharge characterization 7.2.8.1 Histogram formulation Temporal aspects o f spike discharge are typically examined in responses to repeated presentations o f the same stimulus. To assess the spike interval distribution, we computed inter-spike interval histograms (ISIHs). The spike distribution was related to the stimulus time in peristimulus time histograms (PSTHs). In both cases, the x axis was "binned" into a number o f periods o f equal size. Histograms were formulated on the basis of spikes' times o f occurrence during the 200 ms voltage response. Spike times were determined by a voltage detector (A/Dvance version 3.0.9, McKe l l a r Designs, Vancouver B C ) . The time of occurrence for each action potential was recorded as the spike crossed 0 mV. Dot raster displays illuminated the timing o f action potentials relative to stimulus onset, for the individual trials. T. J. A D A M 69 Discharge patterns were characterized in P S T H s and ISIHs. Each P S T H was constructed from dot raster displays by dividing the 200 ms stimulus/response period into 0.10 ms bins. Spike occurrences were then pooled in these bins over 50 trials. The ISIHs were also constructed from dot raster displays, by extracting the inter-spike intervals, also with a binwidth of 0.1 ms, to summarize the neuron's discharge rate and regularity. 7.2.8.2 Basic firing statistics Preliminary recordings and previous work conducted in vivo demonstrate that L S O neurons display significant temporal changes in spike discharge over the stimulus period (Tsuchitani, 1982; Young, Robert and Shofner, 1988; Young, Shofner, et a l , 1988). We examined statistical aspects o f firing behaviour (e.g. rate accommodation, timing precision) as functions of time during the stimulus period over repeated trials (50). From the spike latency data used to construct P S T H s and ISIHs, the discharge patterns o f neurons could be characterized by basic descriptive statistics. Means and standard deviations were computed for onset spike latency (first spike in the train), and for each inter-spike interval " i " occurring in the spike train. The number o f action potentials in repeated spike trains varied such that the last one or two were not present in all 50 trials. Therefore, a criterion for inclusion was set such that an inter-spike interval was only included in the analysis i f the spike terminating that interval was present in > 40 o f 50 trials. The mean ( L S i ) and standard deviation ( S D L S i ) for onset spike latency was computed as: T. J. A D A M 70 S D L S , 2 ( L S 1 " J L 8 1 f j - l where j is the number o f trials (50). For repeated discharge, the mean and standard deviation for intervals in the train (corresponding to L S l + i - L S i ) occurring over repeated trials (ISI ; and SDisiO were computed as: _ ( I S I | ) Easii-jisrr IS I ; = 2 SD r a j=i * j » j - l where j is the number of trials (50). For each interval " i " in the spike train, instantaneous spike rate (Rj) was computed as follows, and then normalized to the first interval, and expressed as a percentage (RA;): 1 100* ( R j - R : ) R>=isi- R A . » » = — R T ^ Nonstationarities in discharge rate (spike rate accommodation, S R A ) were then examined over the stimulus period by fitting single or double exponential curves (Igor Pro, Wavemetrics, Inc.) to the time course o f discharge rate changes at each stimulus level. 7.2.8.3 Discharge regularity The chopper firing pattern combines regularity with rate accommodation. The regularity is, therefore, not stationary. To quantify this regularity, we computed the mean and standard deviation of inter-spike intervals as a function i f time into the stimulus (Young, Robert and Shofner, 1988; Young, Shofner, et al., 1988). The precision of action potential timing was also a nonstationary process, and therefore studied as a function of time into the stimulus period. Sequential interval statistics (ISI; and S D ^ i ) were employed in a regularity analysis used previously to T. J. A D A M 71 characterize discharge behaviour in neurons of the cochlear nucleus (Young, et al., 1988a). A s an index for variability in spike timing, we adopted the coefficient o f variation (CVi ; Young, et al., 1988a; Young, et al., 1988b). This index is a more accurate reflection of timing precision than the standard deviation (SD I S I i ) , as it accounts for the correlation of interval variability with interval duration. The C V i normalizes the standard deviation to the mean interval: where " i " is the interval number within the spike train. For analysis o f spike rate accommodation, double exponential curves were chosen over single exponential fits only when significant (p<.05) decreases in error o f the curve fit were obtained over single fits (Igor Wavemetrics) by performing F-tests on chi-square (X ) values: (x 2 s ing le -x 2 double ) (X s ingle / (n-6) ) Group differences between neuron classes were tested with the two-tailed nonpaired samples Student's t-test. Quantitative pharmacological effects were tested for significance using the two-tailed paired samples Student's t-test as well. 7.2.9 Neurobiotin injection and related histology A subset o f L S O neurons was morphologically identified by intracellular marking with neurobiotin through whole cell patch electrodes containing 0.5% N-(2-aminoethyl) biotinamide hydrochloride, dissolved in the standard internal solution (Vector Laboratories, Burlington, ON) . Following the recording session, the tissue slice was removed from the recording chamber and fixed (4% paraformaldehyde in 0.1 M sodium phosphate buffer solution with 20% sucrose for cryoprotection) for 1 to 4 days. Slices were resectioned (100 um) using a freezing microtome. Sections were then processed for T. J. A D A M 72 neurobiotin histology (Ki ta and Armstrong, 1991) by incubation in the secondary antibody of the avidin-biotin-HRP complex for 2 hours ( A B C kit, Vector Labs Canada, Inc.). The H R P labeling was intensified using the di-aminobenzidine nickel/cobalt method. Sections were dehydrated, cleared, and mounted on chrom-alum slides, left to dry for twelve hours and then coverslipped. Reconstruction of stained neurons was achieved from adjacent tissue sections using camera lucida drawings. The boundaries o f the L S O were readily visible, without counterstaining. 7.3 Results Neurons were only included in the database i f they exhibited resting potentials (0 n A constant current injection) o f -55 m V or less, and overshooting action potentials exceeding 60 m V in height. These criteria are similar to those employed by W u (1991) for in vitro preparations o f the auditory brainstem. A total o f 249 L S O neurons were included in the present tissue slice investigations. The electrophysiological properties and morphology of L S O principal neurons and efferent neurons of the lateral olivocochlear system have been published previously (Adam, Schwarz, et al., 1997; Schwarz, Tennigkeit, et al., 1997). 7.3.1 Two electrophysiological types of LSO neuron are distinguished by morphology T w o electrophysiological cell types were observed in tissue slice recordings from the L S O . Both classes were observed to coexist in the L S O , being observed at all ages studied, in all limbs of the L S O , and frequently in the same tissue slice. Previous research delineating the response properties and cytoarchitecture o f L S O neurons suggests that principal neurons projecting to the inferior colliculus and descending efferent neurons of T. J. A D A M 73 the lateral olivocochlear system are both located in the L S O . We set out to distinguish these neuron types on the basis of their intrinsic membrane properties. We labeled the neurons during whole cell patch recordings with electrodes containing neurobiotin. After electrophysiological properties and spike discharge were delineated, the tissue slices were fixed, and the cellular morphology revealed through histological processing for the neurobiotin marker (see Methods). The soma and dendritic arborization o f each neuron were then reconstructed from adjacent transverse sections in camera lucida drawings. A l l intracellularly stained neurons were localized within the L S O . A direct correspondence between spike discharge patterns and cellular morphology was evident. A majority o f L S O neurons (n=178, 75 %) responded to depolarizing current pulses with transient repetitive firing at a short response latency. This was reminiscent of chopper discharge described in anaesthetized in vivo preparations. Consequently, we referred to these neurons as "chopper neurons". A minority (n=58, 25 %) elicited repetitive spike discharge upon depolarization, with a pronounced and variable delay in onset during moderate stimulation. These neurons were therefore termed "delay neurons". Basic membrane properties also distinguished chopper neurons and delay neurons. For ease of comparison, these are presented in Table 5.02 (intracellular recordings, 34 °C) and Table 5.03 (whole cell recordings, 22 °C). In intracellular recordings, chopper neurons showed resting potentials averaging-61.2 ± 4 . 8 m y , input resistances of 28.1 ± 8.0 mQ, and onset membrane time constants of 2.3 ± 0.9 ms, measured during small (< 5 mV) hyperpolarizations. Ac t ion potentials were 0.6 ± 0.2 ms wide at half their height from the spike inflection point, and rose 64.9 ± 6.2 m V from rest. The resting potentials o f delay neurons were similar to chopper neurons, averaging-61.7 ± 5.4 m V (n=34). However, the input resistances and onset membrane time constants for delay neurons were significantly higher than observed in chopper neurons (60.8 ± 16.8 M Q , and 5.5 ± 3.2 ms, respectively, n=34). Act ion potential amplitudes were 70.9 ± 8.0 m V tall, and 0.7 ± 0.2 T. J. A D A M 74 ms wide (spike half-width). The two electrophysiological cell types were therefore distinguishable on the basis of their respective membrane properties, as well as spike discharge. Morphological identification of chopper neurons and delay neurons revealed that they correspond, respectively, to principal neurons and lateral olivocochlear efferents residing in the L S O . Chopper neurons (n=13) possessed the general morphological characteristics o f L S O principal neurons (Helfert and Schwartz, 1986). In transverse sections, their somata were large and fusiform, averaging 27 x 14 u M in diameter. T w o to three principal dendrites extended from the longer somatic axis in a bipolar fashion, and branched sparingly into several smooth processes. The arborization extended to the margins o f the L S O , along the isofrequency axis (perpendicular to the nucleus curvature). A typical representation o f L S O principal neuron morphology is depicted in Figure 5.02 A , with its location in the L S O . The axons of L S O principal neurons marked could not be traced outside of the L S O margins (not shown). Morphological features of stained L O C efferents (n=3) were distinct from L S O principal neurons, although they were intermingled with L S O principal neurons within the confines o f the L S O (Figure 5.02 B ) . The somata o f L O C efferents were smaller (~ 15 u M in diameter), and approximately spherical. T w o to three principal dendrites extended from the soma, and divided into numerous fine processes which had a highly tortuous appearance. Unlike L S O principal neurons, the dendritic arbors of L O C efferents did not appear to be confined to the isofrequency laminae of the L S O . In addition, the orientation o f the L O C efferent dendritic arbor varied with the cell's location within the L S O . Two neurons marked in the middle and lateral limbs o f the nucleus had arborizations oriented predominantly perpendicular to the nucleus curvature. However, in the medial limb, the arbor was nearly parallel to the curvature of the L S O (Figure 5.02 B) . The axons of L O C efferents coursed mediodorsally upon leaving the L S O (not shown). These data suggest a correspondence o f chopper-like spike discharge to L S O principal neurons and delayed repetitive firing to efferents of the lateral olivocochlear T . J . A D A M 75 system. N o other cell types were observed in tissue slice recordings and there was no ambiguity in the classification. 7.3.2 Action potential discharge in LSO principal neurons A primary goal of the in vitro experiments was to identify possible intrinsic mechanisms for L S O chopper response generation. This was considered to be critical, as it is difficult to relate to those o f either its inputs, or the stimulus waveform. This observation led to our hypothesis that chopper responses are produced chiefly by intrinsic membrane properties of the L S O principal neuron, which then interact with synaptic drive to yield the reported chopper response dynamics. To this end, the spike discharge behaviour of L S O principal neurons in the tissue slice was examined during direct depolarization of the membrane potential. A range of current pulse amplitudes was employed to examine spike discharge over a range of depolarization levels in the tissue slice. Principal neurons o f the L S O responded to increasing depolarization amplitudes with increases in firing rate, duration, and spike count. During peri-threshold stimulation, L S O principal neurons exhibited a single onset spike at a short latency of 2 to 3 ms from stimulus onset. Greater depolarizations evoked transient repetitive spike discharge (Figure 5.03), with a rate and duration that increased with current amplitude. Firing was rapid early in the response period, slowed subsequently and terminated prior to stimulus offset. Only very strong current pulses (e.g. 1 nA) evoked firing that persisted for the duration of the 200 ms stimulus period, and this typically compromised neuronal viability. Thus, a hypothetical rate code o f IID seems to be incompatible with the discharge pattern of L S O principal neurons. These observations already reveal parallels in spike discharge o f L S O principal neurons in the tissue slice during direct depolarization and in the anaesthetized preparation during tonal stimulation. Act ion potential discharge is early, transient, T. J. A D A M 76 repetitive, and exhibits short-term adaptation. This suggests that the phenomena observed in vivo, that have been so difficult to relate to stimulus parameters and input discharge patterns, are largely a manifestation o f the membrane properties o f the L S O principal neuron. For example, the apparent increases in spike count, discharge rate and duration associated with depolarization amplitude are analogous to changes observed in the previous chapter, during monaural ipsilateral auditory stimulation o f EE units. This suggests that depolarizing current pulses used in vitro may correspond, to an unexpected extent, to ipsilateral excitation in vivo. A n exciting possibility is that the complex interactions between ipsilateral excitation and contralateral inhibition in L S O chopper neurons observed previously may also be largely rooted in the intrinsic membrane properties o f these neurons. This issue is revisited in the following sections. 7.3.3 Chopper-like firing patterns in the tissue slice Demonstration o f intrinsic chopper response generation mechanisms requires the satisfaction of criteria outlined in previous reports defining the chopper response. These include (i) a precise onset latency, (ii) regular repetitive discharge, and (iii) spike rate accommodation (e.g. Young, et a l , 1988a; Tsuchitani, 1982). We replicated the response parameters, at least qualitatively, in the tissue slice for L S O chopper neurons stimulated with direct depolarizing current injection. Responses to 200 ms current pulses were recorded and P S T H s constructed from 50 repetitions o f each depolarization. Firing was then characterized with respect to the onset latency precision, regularity o f repetitive firing, and spike rate accommodation. In the tissue slice, chopper-like responses were evident in the multiple peaks of spike latency distributions. These multi-modal P S T H s reflected regular repetitive discharge during a spike train that was firmly locked in time to the stimulus onset (Figure 5.03 A ) . The first mode was narrow, with the onset spike typically falling within one or two 0.1 ms bins (Figure 5.03 A , arrows). This indicated that the onset of the spike train T. J. A D A M 77 was strictly locked to the stimulus onset. Subsequent modes reflected the precise timing of repetitive firing. These peaks gradually became lower and broader, reflecting greater dispersion o f spike timing, until modes were eventually obscured. Increased spread of modes during the stimulus reflected a cumulative increase in spike latency variability. This was apparent in a dot raster display o f spike latencies contributing to the P S T H , where early timing errors were carried through to later action potentials in each response (Figure 5.03 B) . Spike latency distributions, therefore, indicate that for spike trains evoked during direct depolarization of the L S O principal neuron, the timing o f individual action potentials is highly regular, relative to the onset o f the stimulus. This represents chopper-like firing in the tissue slice. Chopper-like spike discharge was also evident in inter-spike interval histograms constructed from responses (Figure 5.03 C). Tight clustering o f inter-spike intervals around the mean resulted in narrow uni-modal interval distributions. This is indicative of action potentials that are precisely timed with respect to each other. Therefore, ISIHs demonstrated that repetitive firing in L S O principal neurons was regular. Additionally, interval distributions exhibited modest positive skew, produced by longer inter-spike intervals. Demarcation of intervals occurring late in the spike train (grey bars) indicated that inter-spike intervals were gradually lengthened as the spike train progressed (Figure 5.03 B) . This is directly reflective of spike rate accommodation in L S O principal neurons. Together, the narrow distribution and modest positive skew o f inter-spike interval distributions supported stimulus-locked discharge onsets, regular repetitive discharge, and spike rate accommodation in L S O principal neurons. A n interesting feature of chopper responses in the anaesthetized preparation is the relative independence of timing precision from stimulus intensity. Unlike other neurons, chopper discharge timing is precise over the range from peri-threshold stimulation to maximal firing (Tsuchitani, 1982; Young, Robert and Shofner, 1988; Young Shofner and Robert, 1988; Holt , Softky, et al., 1996). This was also evident in the chopper-like discharge observed in the tissue slice. Spike discharge produced multi-modal T. J. A D A M 78 PSTHs , and narrow positively skewed ISIHs at all suprathreshold depolarization levels (cf. P S T H s and ISIHs for different depolarization levels in Figure 5.03 A , B ) . Consequently, the pronounced spike timing precision so unique to the chopper response is, at least, supported by intrinsic membrane properties. Clearly, mechanisms contributing to the generation o f the chopper response pattern reside in the membrane of the L S O principal neuron. 7.3.4 Discharge statistics for chopper-like patterns in the tissue slice A n estimate of the role of intrinsic membrane properties in chopper-like firing requires a quantitative description of spike discharge over the stimulus period and over a range o f stimulus intensities. Discharge statistics were computed to obtain quantitative descriptors of spike timing and regularity as a function o f depolarization amplitude. The stimulus dependency o f discharge onset, rate, and regularity were then compared to our previous in vivo investigation of short-term adaptation in IE units. Analogies drawn between these experiments suggest, indeed, a decisive contribution of intrinsic membrane mechanisms to the generation of the chopper response pattern. Discharge statistics were computed for each spike latency and inter-spike interval occurring in a train over repeated (50) presentations o f a constant amplitude current pulse. Direct measures included the mean and standard deviation of onset spike latency, as well as the mean and standard deviation o f each inter-spike interval occurring in a train. Derived measures included the coefficient o f variation (CV;) for each inter-spike interval in the train and rate accommodation (see Methods). A l l indices are depicted graphically as a function o f time into the stimulus period and as a function of depolarization amplitude in Figure 5.04. Onset spike latency was exceptionally short in the chopper-like discharge (< 3 ms). A s indicated i n Figure 5.04 A (inset), the onset latency was brief at all stimulus levels, from near threshold (2.5 ms) to maximal firing (1 ms). With increased current T. J. A D A M 79 amplitude, onset latency decreased monotonically, approaching an asymptotic value o f 1 ms. Therefore, onset latency typically decreased only minimally over the dynamic range of voltage responses recorded (< 2 ms). Timing precision was highest for the onset spike of the chopper response. The standard deviation {error bars) o f onset spike latency was less than 5 % of the mean at peri-threshold levels, and decreased further with stimulus amplitude. Properties o f repetitive discharge were characterized as changes in inter-spike interval with latency, for each spike in a train, from stimulus onset and with current amplitude. A s the spike train proceeded, successive inter-spike intervals were lengthened (Figure 5.04 A ) , reflecting spike rate accommodation. A s the magnitude of depolarizations was increased, overall inter-spike intervals were shortened, producing a reduced slope in the relation o f inter-spike interval to stimulus amplitude (Figure 5.04 A ) . Thus, spike rate in chopper-like responses increased with stimulus strength and decreased with time into the stimulus period. Spike timing precision during repetitive firing also decreased with time from stimulus onset, and increased with current amplitude. Error bars reflecting the standard deviation o f inter-spike interval durations increased with time into the stimulus period (Figure 5.04 A , error bars), and decreased with stimulus intensity (Figure 5.04 A , cf. error bars <across plots). Therefore, timing variability did appear to accumulate during the stimulus period and was smaller, overall, for stronger stimulation. However, the variability of inter-spike intervals was correlated to their mean duration. A s depicted in Figure 5.04 B , the CV; was low, and fairly stable throughout the repetitive discharge and across stimulus level. A t all suprathreshold stimulus levels, CVj was smaller than 0.1, and increased only slightly over the duration of the response, remaining below 0.2 for all intervals in the train. Therefore, the increases in inter-spike interval during the response (Figure 5.04 A ) did not compromise the relative spike timing precision (regularity) in the chopper-like response. The CV; was also inversely related to T. J. A D A M 80 current pulse amplitude. Chopper-like discharge in the tissue slice is, consequently, characterized by a high degree of regularity throughout the response. This, too, parallels findings reported in the anaesthetized in vivo preparations (Tsuchitani and Johnson, 1985; Tsuchitani, 1982; Young, Robert and Shofner, 1988; Young, Shofner and Robert, 1988). Chopper-like firing patterns are distinctive in their superior spike timing precision, particularly for the onset spike, even at peri-threshold levels. A s is evident in Figure 5.04 A , inter-spike intervals increased during the stimulus period, towards an asymptote. This directly reflected spike rate accommodation in chopper-like discharge. To compare rate accommodation across stimulus intensity, instantaneous spike rate was normalized to the first inter-spike interval and plotted against time. For all chopper-like responses, spike rate accommodation reduced instantaneous firing rate by approximately 55 to 70 %, prior to the cessation of firing. This occurred over time courses that were optimally characterized by single exponential fits with time constants of 15 to 27 ms in 16 of 23 chopper neurons. Significantly better fits were obtained in the remaining neurons using double exponentials yielding time constants o f 3 to 8 ms, and 32 to 97 ms (n=7). However for the entire sample, normalized accommodation was not related to current pulse amplitude, the exponential time courses remaining stable over all spike trains recorded for each neuron (Figure 5.04 C) . Therefore, chopper-like discharge in tissue slices exhibited spike rate accommodation, which was independent of stimulus amplitude when normalized to the first inter-spike interval. Finally, spike discharge level was examined in chopper-like discharge patterns as a function o f current pulse amplitude. The average overall spike rate, the average spike rate during the first 25 ms of the spike train, and the total spike count were measured as functions o f stimulus strength. O f these, spike count and peak discharge rate were significantly and monotonically related to current pulse amplitude. The increase in rate was approximately linear, with an average slope of 264 ± 23 spikes/second/nA. Complexities in relating discharge level to stimulus amplitude in the tissue slice parallel those observed during our in vivo experiments delineating short-term adaptation in the T. J. A D A M 81 L S O during monaural excitation. Evidently, spike count, and peak and sustained discharge rates are all able to carry information regarding stimulus intensity during depolarization. 7.3.4.1 Relation o f chopper-like discharge to membrane properties The unique chopper response pattern observed during current pulses in vitro must be imposed, most likely, by membrane properties o f L S O principal neurons. A s mentioned previously, L S O principal neurons exhibited a single onset spike at a short latency o f 2 to 3 ms from stimulus onset, during peri-threshold stimulation. This corresponded to the first mode of the P S T H (Figure 5.03 A , arrows). During repeated peri-threshold stimulation with a constant pulse amplitude, the membrane potential frequently failed to reach spike threshold. Comparison o f voltage responses with and without the onset spike revealed that the onset action potential was superimposed on the peak of a transient depolarizing potential (TP) occurring at response onset (Figure 5.05 A , inset). The temporal correspondence between subthreshold and suprathreshold events (TP and onset spike) suggested that the T P triggers the onset spike, producing the short and precise latencies exhibited by chopper-like responses. Greater depolarizations evoked transient repetitive firing (Figure 5.05 A ) . Spike trains were initiated early in the voltage response, exhibited spike rate accommodation, and terminated prior to pulse offset. Both the firing rate and duration of chopper responses increased with current amplitude. Repetitive firing appeared to require greater depolarization to overcome the hyperpolarizing and/or shunting effect of a sustained outward rectification observed in subthreshold responses (see below). This rectification may contribute to excitatory short-term adaptation observed in vivo. Chopper-like responses were also influenced by faster membrane non-linearities, evident as small damped voltage oscillations between and following action potential discharge (Figure 5.05 A , arrows). Depolarizing phases of these oscillations coincided with expected spike latencies when a spike failed in a train o f action potentials, or upon T. J. A D A M 82 termination of the train. These subthreshold voltage oscillations may contribute to the distinctive regularity o f spike discharge in chopper neurons. Subthreshold membrane behaviour of L S O principal neurons was characterized by matching voltage traces for equal magnitude depolarizations and hyperpolarizations. Around rest, voltage trajectories during depolarizations were mirror images of responses to hyperpolarizations. During negative current pulses, the membrane potential was initially hyperpolarized to an early peak. The membrane voltage then relaxed to a steady and relatively depolarized level for the remainder of the stimulus (Figure 5.05 B , arrow). The depolarizing voltage sag commenced approximately 15 ms into the stimulus, and was visible in responses traversing voltages 5 m V or more below rest. A t the offset of the negative current step, the membrane potential exhibited a small (< 10 mV) , brief (< 20 ms) depolarizing afterpotential ( D A P ) , which increased in amplitude with greater hyperpolarizations (Figure 5.05 B , filled rectangle). The D A P did not exceed spike threshold. Subthreshold positive current pulses produced, similarly, a peak-sag-afterpolarization profile in voltage responses. A depolarizing transient potential (TP) was evoked near response onset. The amplitude o f the T P grew with increasing positive current injection, and the latency to peak o f the T P shortened from approximately 10 ms to 2 ms just below spike threshold (Figure 5.05 B , star). The T P was followed by a repolarization towards rest which persisted for the remainder of the stimulus period. A t pulse offset, the membrane voltage trajectory returned to rest after a small (< 10 mV) hyperpolarizing afterpotential ( H A P ) o f less than 25 ms duration (Figure 5.05 B , open rectangle). The H A P also grew with depolarizing current, as expected from the gradual decay o f a sustained outward rectification. The current-voltage ( T V ) relations constructed from voltage responses underscored the balance in time- and voltage dependence o f subthreshold L S O principal neuron membrane behaviour. In all cases, I -V relations for voltage relations evoked from rest were near linear; depolarizations and hyperpolarizations being matched for both peak T. J. A D A M 83 and sustained responses. However, decreased slope resistances were evident for steady state I -V relations (Figure 5.05 C , arrows). Further, the decay of voltage responses at current pulse offset was accelerated relative to onset trajectories. This suggests faster membrane time constants, due to lower resistances, at the offsets of both de- and hyperpolarizing pulses. In any case, underlying conductances are balanced during depolarizations and hyperpolarizations. They also follow similar time courses. This results in the mirror-image of voltage responses (Figure 5.05 B ) , and linear I - V relations (Figure 5.05 C ) for both early (at 10 ms) and steady-state (at 195 ms) responses. Significant and balanced contributions of time- and voltage dependent rectifications in the depolarized and hyperpolarized voltage ranges determine the subthreshold behaviour o f the L S O principal neuron. Such rectifications produce relaxations in voltage responses to both positive and negative current pulses. Their balanced effects yield near linear I -V relations over all time periods of the voltage response, with decreased slope resistances for steady state responses. This subthreshold behaviour appears to influence chopper-like discharge during depolarizations by producing an emphasis on response onset in voltage responses. A T P evidently contributes to the occurrence and timing o f the strictly time-locked onset action potential, and a subsequent hyperpolarizing voltage sag and conductance shunt contributes to spike rate accommodation and termination of firing. Therefore, several of the features o f chopper-like discharge patterns are shaped by membrane rectification in the L S O principal neuron. Particularly by generation o f the T P , membrane properties contribute to the emphasis o f response onset and the fixed latency that is so critical for chopper responses recorded in vitro or in vivo. 7.3.5 Voltage dependence of membrane rectification in LSO principal neurons The previous intracellular recordings indicated that the chopper response, as generated by direct current depolarization in vitro, is a product of intrinsic membrane T. J. A D A M 84 properties of L S O principal neurons. These properties were investigated further with whole cell patch clamp recordings in tissue slices (n=T02). This recording configuration facilitated experimental manipulation of membrane potential from which depolarizations could be evoked. Voltage dependence of membrane behaviour in L S O principal neurons guided subsequent pharmacological experiments dehneating underlying ionic membrane conductances, and suggested hypotheses as to their contributions to chopper behaviour during tonal stimulation. The chopper-like behaviour of L S O principal neurons showed strong voltage dependence, as revealed in responses evoked from various membrane potentials (holding potentials). These were imposed by constant current injection through the recording electrode, and current pulses were evoked from this baseline. Spike discharge as well as subthreshold membrane properties were significantly altered by manipulation o f the holding potential (n=T02). Linear current-voltage (I-V) relations derived from voltage responses evoked at rest (0 m V constant current) belied significant membrane non-linearities in the subthreshold behaviour o f L S O principal neurons. Further, the voltage dependence of these membrane non-linearities impacted the subthreshold behaviour o f the L S O principal neuron, as well as spike discharge. Voltage dependence of the primary features of L S O principal neuron behaviour is considered below. 7.3.5.1 Chopper-like firing ;> The possible voltage dependence o f chopper-like discharge in tissue slices is particularly interesting considering excitatory and inhibitory interactions in the L S O principal neuron. Determining the voltage dependence of chopper neuron behaviour would be helpful in the delineation of possible inhibitory effects on chopper responses. O f course, inhibition occurs via glycinergic conductances not manipulated in our tissue slice experiments, but the influence o f converging inhibition on ipsilaterally evoked chopper responses wi l l in part be explained by their voltage dependence. A s well , the T. J. A D A M 85 voltage dependence o f subthreshold membrane properties contributing to chopper responses wi l l guide the pharmacological delineation of underlying membrane conductances. Voltage dependent rectification o f the chopper membrane wi l l in part determine the behaviour o f the L S O principal neuron during ongoing stimulation in natural listening. Chopper-like responses were compared for depolarizations evoked from different holding potentials, where steady state voltage levels were matched (Figure 5.06 A , star). The chopper discharge pattern was influenced by holding potential (membrane potential at the onset of a depolarizing current pulse. The onset spike latency was relatively invariant during even large shifts in holding potential (-15 to +10 m V shifts). However, repetitive firing was dramatically affected in Figure 5.06 A . The depolarization from -62 m V evoked two action potentials within 15 ms. Repetitive firing of longer duration required larger current pulses. During pre-hyperpolarization below ~ -65 m V , only onset action potentials could be elicited, regardless of pulse amplitude (Figure 5.06 A , -74 mV). Imposing holding potential shifts positive to rest resulted in a decreased discharge rate of repetitive firing probably due to the activation of shunt conductances in the depolarized range (Figures 5.07, 5.09). A t more positive holding potentials, repetitive discharge was eliminated, leaving only an onset spike. In Figure 5.06 B , the current pulse injected from rest (-62 mV) was reduced to a subthreshold amplitude, yielding, after the T P , a steady-state potential (*) that reached a lower voltage level than in A . A current pulse leading to the same steady-state depolarization from a hyperpolarized holding potential produced a much faster rise o f the T P that consequently triggered an action potential. The T P (without the onset spike) was nearly abolished when the same steady-state voltage level was achieved from a depolarized holding potential. Chopper-like responses yielding the same net (absolute) depolarization (i.e. a small pulse evoked from a depolarized holding potential versus a large pulse from a hyperpolarized holding potential) do not, therefore, yield the same spike discharge in T. J . A D A M 86 LSO principal neurons. In the tissue slice, chopper-like response magnitude does not directly reflect depolarization amplitude, as it also varies with the membrane potential from which a depolarization is evoked. Conversely, the onset timing of the chopper response is largely conserved over the entire range of holding potentials tested (-50 to -75 mV). This is largely due to an enhanced TP for hyperpolarized holding potentials, in spite of a pronounced voltage dependence of the TP peak latency in the subthreshold range (Figure 5.07 B, C). These observations illustrate that hyperpolarization predisposes LSO chopper neurons to emphasize the onset of the response. Thus, the emphasis of response onset during binaural stimulation in vivo is well explained by intrinsic membrane properties, and does not require an assumption of earlier arrival of EPSPs. 7.3.5.2 Membrane rectification in the depolarized voltage range Subthreshold responses of LSO principal neurons exhibited voltage dependence that complemented that of spike discharge. The TP certainly contributes to the onset response. Its voltage dependence is shown, to some detail, in Figure 5.07. Hyperpolarization of the holding potential increased the voltage range to be traversed to spike threshold, and the relative prominence of the TP in depolarizing voltage responses. Further, the TP arose from a distinct inflection point when evoked from negative holding potentials (Figure 5.06 B, Figure 5.07 B-C). This suggests that depolarizing membrane conductance(s) contributed to the TP. Activation of this conductance appeared to be voltage dependent, as evident in I-V relations derived from voltage responses (Figure 5.07 D-F, below respective recordings). The transient inward rectification, apparent in plots for each holding potential (small arrows), was more evident in relations obtained from negative holding potentials, where a positive deviation from linearity reflected the TP (Figure 5.07 E-F, small arrows). There was an apparent voltage threshold for the TP and the rectifying conductance at approximately -65 mV (n=20), as indicated by the deviation T. J. ADAM 87 from linearity in the depolarizing direction (Figure 5 . 0 7 F). The apparent inward membrane conductance(s) that supports the rise of the TP promotes the occurrence of the onset action potential within a narrow time window and seems critical, therefore, in the generation of the chopper response. A hyperpolarizing voltage sag may also shape the TP by sculpting its decay phase. Current-voltage relations indicated an outward rectification in the depolarizing voltage range (Figure 5 . 0 7 D-F). Comparison of steady state I-V relations evoked from hyperpolarized holding potentials demonstrated that this rectification was not dependent on holding potential, but was activated during depolarizations above - 5 0 mV (Figure 5 . 0 7 B). This is consistent with the contribution of a voltage dependent outward conductance which, likely, contributes to spike rate accommodation and chopper response termination. 7.3.5.3 Rectification in the hyperpolarized voltage range Changes in chopper responses during natural listening may also be influenced by the voltage dependence of subthreshold membrane properties in the hyperpolarized voltage range. Furthermore, these properties may affect chopper responses evoked during depolarization from negative holding potentials through delayed deactivation of underlying conductances. This is revealed by the occurrence of depolarizing afterpotentials (DAPs) following negative current pulse offsets. Rectifications in the hyperpolarized voltage range were delineated in voltage responses evoked by negative current pulses from variant holding potentials. Chopper neurons expressed a depolarizing voltage sag during hyperpolarizations, that reached steady state within 1 0 0 - 1 5 0 ms of stimulus onset (Figure 5 . 0 7 A-B). From I-V relations, this sag was apparent in steady state relations by a depolarizing deviation from linearity below - 6 0 mV (Figure 5 . 0 7 D, arrowheads). Steady state I-V relations possessed lower slope resistances below this membrane voltage, suggesting the activation T. J . A D A M 88 of an inward (anomalous) rectification. The depolarizing voltage sag showed voltage dependence. Negative current pulses evoked less prominent sags when the holding potential was hyperpolarized (Figure 5.07 C) . Comparison of the slope conductances for peak and steady state responses indicated that the sag conductance activated below -60 mV, and approached maximal activation at approximately -70 to -80 m V , since hyperpolarizations from such holding potentials evoked little or no sags. Consequently, the depolarizing voltage sag is likely produced by an anomalous rectification activated on hyperpolarization of the membrane potential below -60 mV. Further, due to its voltage dependence, the underlying conductance probably contributes to the membrane resting potential (0 n A holding current) o f L S O principal neurons. Early hyperpolarizations (peak responses) also showed some non-linearities at very negative potentials (Figure 5.07 A ) . The associated increase in slope conductance had a different voltage range from that for steady state voltage responses. Instead, peak I-V relations showed an inward rectification below approximately -80 m V for this cell. This suggested that a different inward conductance, possessing rapid activation kinetics, also contributed to hyperpolarizing responses. However, the activation voltage for the rapid anomalous rectification varied from cell to cell, within the range of -65 to -85 mV. The delayed inward rectification evident in the depolarizing voltage sag is caused, in most neurons, by a hyperpolarization-activated cation current (LH - Pape, 1996; McCormick and Pape, 1990; Osmanovic and Shefher, 1987; Spain, Schwindt and Cr i l l , 1987). Due to the slow inactivation kinetics o f this current, the sag should be followed, on repolarization, by a depolarizing afterpotential ( D A P ) . Such events, were, indeed, observed in 102 neurons, whenever a sag occurred during the preceding hyperpolarization. Examples are shown in Figures 5.07 to 5.11. Generally, the amplitude o f the D A P was proportional to the depth o f the voltage sag, and its decay time course reflected, roughly, that of the sag (e.g. Figure 5.08 b, middle and right records). Hyperpolarizing responses without the sag were often not followed by a D A P (e.g. Figure 5.08 A , right records). T. J. A D A M 89 Amplitudes and time courses of D A P s could not always be interpreted as gradually deactivating I H ' s , however. A s a rule, D A P s with large amplitudes (> 10 mV) and fast decay time courses appeared to be superimposed on the slower, IH-related events on return from relatively large hyperpolarizations to relatively depolarized membrane potentials (Figure 5.08 A , left and middle records; Figure 5.08 B , left records). Such fast D A P s were also observed in the absence o f pronounced sags. A n inflection in the rising phase of a D A P following a moderate repolarization step (Figure 5.08 A , middle records) hints at a separate inward transient conductance that contributes to the D A P when the voltage step returns to depolarized holding potentials (at rest and above) from deep hyperpolarizations. Below, it w i l l be shown that a transient low-threshold calcium conductance can account for these D A P s in L S O principal neurons. When the holding potential was sufficiently close to spike threshold, the fast D A P s were able to trigger the firing of a single action potential (Figure 5.08 B , left records). This observation further points to the ability o f L S O principal neurons to emphasize the onset of an excitatory response when the membrane has been pre-hyperpolarized (via inhibition). 7.3.5.4 The linearity of current-voltage relations Principal neurons of the L S O have been reported to display linearity in their I -V relations (Wu and K e l l y , 1993; W u and K e l l y , 1991). The measurements discussed above illustrate voltage dependencies of membrane behaviour that strongly affect the output of these neurons. Thus, the apparent linearity o f W u and K e l l y does not predict the input-output functions o f these neurons. Current-voltage relations with nearly linear appearance were also observed in the present study, provided measurements were restricted to the steady-state voltage responses to current pulses injected from levels near rest (no holding current, e.g. Figure 5.07 E) . A reason for this quasi-linear appearance is that the inward and outward T. J. A D A M 90 rectifications in the hyper- and depolarized voltage ranges can be o f similar magnitudes. The non-linearities become more apparent when the I -V functions are measured from different holding potentials, and when the pronounced time-dependent voltage shifts in both de- and hyperpolarized ranges are taken into account (Figure 5.07 D-F) . In our experience, the depolarizing voltage sag (which was not pronounced in W u and K e l l y ' s records) quickly disappears when recording conditions deteriorate. It is therefore possible to measure different degrees of linearity in the same neurons under different recording conditions. A n important conclusion from the measurements reported here is that I-V relations may appear almost linear, in spite of the contribution of strongly voltage-dependent conductances (see below) that can produce strong time-dependent non-linearities in the output. 7.3.6 Conductances contributing to chopper membrane dynamics Intrinsic membrane properties o f the L S O principal neuron are bestowed by the interplay o f membrane conductances residing in the neuronal membrane. The repertoire of intrinsic conductances molds, therefore, the membrane potential and chopper responses to synaptic or injected current stimulation. Voltage- and time dependence o f membrane properties revealed in responses to current pulses (above) guided the following pharmacological identification of underlying ionic membrane conductances. The identities of resident membrane conductances were delineated in whole cell patch recordings performed on 81 L S O principal neurons in the L S O of the rat auditory brainstem slice preparation. 7.3.6.1 Conductances shaping the depolarizing voltage sag The voltage- and time dependence o f rectification in the hyperpolarized range suggested that L S O principal neurons possess strong anomalous rectification that limits T. J . A D A M 91 the magnitude and duration o f hyperpolarization in voltage responses. Candidates for anomalous rectification included a hyperpolarization-activated cation conductance, and a rapidly activating inwardly rectifying potassium conductance (Pape, 1996; Nichols and Lopatin, 1997). Voltage responses to hyperpolarizing current pulses were recorded during selective blockade of each of these conductances. Anomalous rectifiers were distinguished on the basis of their differential sensitivity to barium (Ba 2 + ) and cesium (Cs + ) . Barium was applied externally to block the rapidly activating inward rectifier, and C s + was added to the A C S F to additionally block the hyperpolarization-activated cation conductance (Pape, 1996; Nichols and Lopatin, 1997). These experiments were conducted in 600 n M T T X , to block the transient sodium conductance ( G N a T ) , which underlies the rising phase o f the action potential in most neural systems. This suppressed spike discharge in L S O principal neurons and synaptic inputs, stabilizing the membrane potential. However, an ancillary effect o f T T X was to hyperpolarize the resting potential by approximately 3 mV. This indicates the presence o f a persistent TTX-sensit ive conductance in L S O principal neurons, and suggests a contribution o f a persistent sodium current to the resting potential. We compensated for the hyperpolarization through the introduction o f a constant offset current to bring the membrane potential to its original level. Voltage responses were then evoked from holding potentials in the range of -55 mV, above the reported voltage activation range for anomalous rectifiers. Barium application (0.2 m M ) depolarized the resting potential by 7.2 ± 2.2 m V (n=7, p<0.01), and significantly increased the input resistance by 61.8 ± 27.9 % ( p O . O l , for hyperpolarizations under 5 mV). This indicated that B a 2 + blocks conductances that contribute to the resting potential, and affect responses to current injection from rest. Barium application most notably amplified hyperpolarizing responses, and revealed a stronger depolarizing voltage sag during negative current pulses (Figure 5.09 A -B , arrows). This indicates the contribution o f a rapidly activating anomalous rectifier T. J. A D A M 92 (GKIR ) to L S O chopper neuron behaviour in the negative voltage range. Peak I -V relations further support this conclusion, as the slope resistance in the presence of B a is greater than in T T X alone, particularly below -60 m V (Figure 5.09 A , D , hyperpolarizations). The presence B a 2 + significantly increased slope resistance by 77.7 ± 23.6 % (p<0.01). However, slope input resistance for steady state I -V relations was less increased by B a (18 ± 7.8 %), and over a narrower range, positive to ~ -85 mV. This may be due to the shunting effect o f a relatively greater depolarizing sag conductance ( G H ) that has become unmasked at very negative potentials by B a blockade of the rapid anomalous rectifier (Figure 5.09 B , E). A significant contribution o f G K I R to L S O principal neuron behaviour is indicated by both its B a sensitivity and its voltage range. The B a effects below -60 m V are consistent with a major contribution o f G K I R to hyperpolarizing responses of the L S O principal neuron. During depolarizations, slightly increased voltage deflections were evident for 2_|_ peak and steady-state responses after application of B a (Figure 5.09 A - B ) . This was also apparent in peak I -V relations (Figure 5.09 D) . Because of the voltage range of this observation (>-55 mV), and the known multiplicity o f membrane conductances affected by B a 2 + , G K I R is not a candidate for contributing to the influence of B a 2 + on depolarizations. This effect may be the product of ancillary antagonism o f other potassium channels. Addit ion o f 3 m M C s + (n=6) to the external medium (already containing 600 n M T T X and 0.2 m M B a ) hyperpolarized the L S O principal neuron resting potential by 5.0 ± 2.10 m V from the resting potential obtained under B a 2 + . Cesium also significantly increased the input resistance, measured during small hyperpolarizations from rest, by 170.1 ± 3.3 % over those in the presence o f B a 2 + (p<0.01). Conductances blocked by C s + therefore contribute to the resting potential and play a role in chopper neuron behaviour in this voltage region. During hyperpolarizing pulses, the depolarizing voltage sag was eliminated upon application o f C s + (Figure 5.09 B - C ) . Instead, the membrane time T. J . A D A M 93 constant was greatly increased, indicating a greater input resistance in the presence of C s + (Figure 5.09 B , C) . This effect was also evident in steady-state I - V relations below -60 m V (Figure 5.09 E ) . Below this voltage, L S O principal neurons showed increases in slope input resistance o f 482 ± 176.2 % (p<0.01). Responses to depolarizations were not significantly affected by C s + (Figure 5.09 B - C , E ) . These results support the contribution o f a large delayed hyperpolarization-activated cation conductance ( G H ) to the membrane behaviour o f L S O principal neurons in the hyperpolarized voltage range and to their resting potential. This experimental series supports the hypothesis that two families of conductances underlie anomalous rectification in L S O principal neurons: a rapidly activating, non-inactivating anomalous rectifier ( G K I R ; blocked by B a 2 + ) , and a delayed, hyperpolarization activated, non-inactivating anomalous rectifier ( G H ; blocked by C s + ) . Further, the voltage sag produced by G H was largely masked by G K I R activation in L S O principal neurons. 7.3.6.2 Conductances contributing to the depolarizing afterpotential In the previous experiment, the D A P s were enhanced during B a 2 + application (Figure 5.09 A - B , arrowhead). This may be due to B a 2 + blockade o f GKXR , which reduces the shunting and masking effect of G K r R on the D A P , and/or the passage of B a through transient low threshold calcium channels, producing the rebound event. There may, additionally, be some B a 2 + blockade o f other potassium conductances activated by the return of the membrane potential to positive levels. These possibilities were investigated in the following two series o f experiments centering on the D A P . Firstly, external C a 2 + was reduced to 200 u M in A C S F containing 600 n M T T X to investigate a possible contribution from calcium conductances (n=8). Secondly, N i was applied in addition to T T X , B a 2 + , and C s + to evaluate the specific contribution of low-threshold transient T. J. A D A M 94 calcium conductances (n=5). In both experiments, responses to hyperpolarizing current pulses were evoked from rest. A pronounced D A P visible in the presence of 600 n M T T X was reduced in A C S F 2+ containing the decreased extracellular C a concentration. Although the membrane time constants at the offset of hyperpolarizations were not significantly affected by this manipulation, the magnitudes o f the D A P s expressed as /dV/dt , were greatly reduced (Figure 5.10). This provides an initial indication that the D A P may be mediated, in part, by a calcium conductance. 2+ The decreased D A P remaining upon Ca reduction was hypothesized to arise from the slow deactivation o f the slowly activating anomalous rectifier ( G H - McCormick and Pape, 1990; Strohmann, Schwarz and Pui l , 1994; Tennigkeit, Schwarz and Pui l , 1996; Huguenard, 1996; Umemiya and Berger, 1994). To test this, B a 2 + and C s + were again co-applied to the external medium containing 600 n M T T X , and hyperpolarizations evoked 2+ from rest (Figure 5.11). The D A P was amplified in the presence o f B a , and occurred at a short and relatively fixed latency from current step offset. In C s + however, the D A P was reduced, its latency prolonged and more variable, probably as a result of the prolonged membrane time constant (Figure 5.11). This supports a contribution by G H . Co-* 2+ application o f N i virtually eliminated the remaining component o f the D A P , supporting the contribution o f a low threshold, transient, calcium conductance ( G C a T ) that was de-inactivated by the hyperpolarizations, and reactivated by its offset. These results indicate that the D A P is supported by more than one conductance, as originally suggested by its complex voltage dependence (see above). Contributing components are G H and G C a T whose interplay determines the D A P ' s voltage dependent dynamics. T. J . A D A M 95 7.3.6.3 Conductances supporting the transient potential Depolarization activated inward currents involving sodium and calcium ions were hypothesized to contribute to the T P in positive voltage responses. Potential candidates included subthreshold sodium conductances, low-threshold calcium transient calcium conductances and the hyperpolarization activated cation conductance. These were examined in a series of experiments involving external calcium reduction, and application of T T X , nickel ( N i 2 + ) , and cesium (Cs + ) . Possible contribution of a subthreshold sodium conductance to the T P was investigated by applying 300-600 n M T T X externally (n=14). Tetrodotoxin effectively blocked sodium action potentials (Figure 5.12 A - B , E-F) in a partially reversible manner, and produced a 3 ± 0.4 m V hyperpolarization of the resting potential. Consequently, a subthreshold sodium conductance contributes to the resting behaviour of the L S O principal neuron. Hyperpolarizing responses were not affected by T T X from any holding potential tested (~ -50 to ~ -80 mV). However, peak depolarizations were decreased under T T X (Figure 5.12 A - B , early voltage response; cf. Figure 5.12 E - F , TP). Tetrodotoxin significantly decreased, but did not eliminate the T P in the range subthreshold to action potential generation. This was also evident in the corresponding I -V relations, where apparent peak slope conductances were decreased by 41.3 ± 10.23 % above -45 m V in the presence o f T T X (Figure 5.12 C , G) , reflecting the blockade o f an inward conductance. In approximately half of L S O principal neurons (n=6/14), steady-state depolarizations were also reduced in T T X (Figure 5.12 A - B , D and E - F , H) . This paradoxical drop in apparent resistance after application of a channel blocker arises because the removal of the depolarizing effect o f the sodium current was greater than the increase o f any voltage response due to the increased resistance. In the remaining 8 out of 14 neurons, both effects may have canceled. Thus, the lack o f change in depolarizing steady-state responses after T T X application does not exclude the presence T. J. A D A M 96 of a persistent sodium conductance which certainly exists in the 6 neurons showing the amplitude decrease. The reduction in the T P and steady-state potentials in T T X indicated the contribution of at least one TTX-sensit ive subthreshold sodium conductance to L S O principal neuron responses during depolarization (herein termed the subthreshold sodium conductance, or G N 3 S ) -The T P amplitude was enhanced when evoked from negative holding potentials (cf. Figure 5.12 B and F). This suggested a contribution from a transient low voltage activated calcium conductance (G C a T)- This possibility was examined by using N i to block G C a T in the presence of 600 n M T T X (n=7, Byer ly and Hagiwara, 1988; Carbone and Lux, 1984; Fox and N o w y c k y and Tsien, 1987; Huguenard, 1996; Umemiya and Berger, 1994). We first hyperpolarized the neurons in order to de-inactivate the Gc a T and then evoked large TPs upon depolarization above -60 m V (Figure 3.14 A ) . Subsequent application of 50 u M N i decreased the amplitude of the T P (Figure 3.14 B) . This was also evident in graphical representation where a the T P amplitude is decreased in the presence o f N i 2 + (Figure 3.14 C). Steady state responses were much less affected by N i 2 + application or not at all (Figure 3.14 A - B , D) . This experiment was subsequently repeated in the absence o f T T X (n=4), to observe possible effects on the onset spike. The 2_|_ application of N i reduced the T P , necessitating greater current pulses to reach threshold (Figure 3.15). Consequently, a G C aT amplifies the T P contributing, thereby, to the short onset spike latency that is so characteristic o f the LSO principal neuron. Finally, it was hypothesized that G H might also contribute to the T P dynamics, since it was found to be a critical contributor to the D A P , in conjunction with a G C a r - We tested this hypothesis in a series o f experiments conducted in the presence of 600 n M T T X , 0.2 m M B a 2 + , 3 m M C s + (to block G H ) and 50 u M N i 2 + in A C S F containing 2 m M C a 2 + (n=6). After T T X application, the T P was evoked upon depolarization from hyperpolarized holding potentials (Figure 5.15 A ) . A more prominent T P was revealed in T. J. A D A M 97 the presence o f B a , that occurred at a short, relatively fixed latency (Figure 5.15 A - B ) . While an increased resistance induced by B a 2 + would amplify any voltage response, the selective enhancement of the TPs could be due, in addition, to the passage of B a through calcium channels. Blockade of the rapid anomalous rectifier was not likely to yield such an effect, because sustained voltage responses were not similarly affected (Figure 5.15 A -B , E ) . Subsequent addition o f Cs increased the onset latency of TPs, which could occur near the end of the 200 ms period at low current amplitudes (Figure 5.15 C). Addit ion of N i 2 + then eliminated the T P (Figure 5.15 C-D) , without further affecting sustained depolarizations. In I -V relations for peak responses (Figure 5.15 C , arrows), the slope (apparent resistance) increased significantly by 180.6 ± 30 % under B a , reflecting increased T P amplitudes (Figure 5.15 F). The elimination of the T P by N i was also represented in I -V relations, where T P amplitude was decreased by 83.1 ± 7.2 % from that in the presence o f C s 2 + (Figure 5.15 E). The effect o f B a 2 + and C s 2 + on the timing of the T P is shown in Figure 5.15 G . The time-to-peak of the T P was significantly delayed only by the application o f C s 2 + . B y blockade of G H , the timing restriction o f the T P was lifted, and the T P behaved in a fashion similar to low threshold calcium spikes described in other neurons, arising from a distinct inflection in the voltage response at a pronounced delay from current pulse onset, and then earlier in the response as pulse amplitude is raised. v The T P is therefore a multi-component event arising from the interplay o f several membrane conductances: the slowly inactivating anomalous rectifier ( G H ) , a subthreshold TTX-sensitive sodium conductance ( G N A S ) , and the low threshold calcium dependent inward conductance ( G C A T ) - Together, these factors produce a prominent depolarizing potential at the onset o f positive voltage responses, which contribute critically to the occurrence and timing of the onset action potential in L S O chopper responses. T. J. A D A M 98 7.3.6.4 Conductances contributing to the hyperpolarizing voltage sag The hyperpolarizing voltage sag terminating the T P may be produced by the interplay of several outward membrane conductances. The possible contributions o f two general classes of depolarization activated potassium conductances to the decay o f the TP , and to the hyperpolarizing voltage sag, were assessed using ascending dosages of 4-aminopyridine (4AP - 50, 200 u M , 1, 4 m M ) and tetraethylammonium chloride ( T E A -0.5, 1.0, 5.0, 20 m M ) in A C S F containing 600 n M T T X . Evaluation o f 4 A P effects on voltage responses during depolarizations was largely precluded in the absence of T T X , due to dramatic spontaneous firing o f L S O principal neurons and possibly their inputs, even at low concentrations of 4 A P . In the presence o f 300 to 600 n M T T X , 4 A P application produced a dosage-dependent depolarization of the resting potential (n=7). Cells used in these studies possessed resting potentials of -62.67 ± 2.73 m V in T T X . Large depolarizations o f the resting potential were obtained at high doses o f 4 A P (1 m M : -42.0 ± 5.7 mV, and 4 m M : -38.5 ± 4.9 mV). Sensitive (4AP) outward conductances are therefore at least partially activated at rest. The effect o f 4 A P on responses to depolarization from hyperpolarized holding potentials was found to depend on the concentration of the agent in the external milieu (n=7). A t low doses (50 to 200 u M ) , 4 A P increased the voltage deflections (resistance) during and immediately following the T P , from both hyperpolarized (Figure 5.16 A ) and depolarized (Figure 5.16 B ) holding potentials. Current-voltage relations plotted for the trough of the transient sag, relative to the membrane voltage in 4 A P over the same time frame revealed increased slope resistances for depolarizations above -55 m V (Figure 5.16 C). Steady-state voltage responses were not significantly affected at low concentrations (50 to 200 u M ) . This is evident in both voltage traces (Figure 5.16 A - B ) and I -V relations (Figure 5.16 D) . A t higher concentrations (1 to 4 m M ) , 4 A P increased steady state T. J . A D A M 99 depolarizations, evoked from either hyperpolarized or depolarized holding potentials (Figure 5.17). Together, these findings suggest the possible coexistence of two 4 A P sensitive potassium conductances differentiable, in part, by 4 A P sensitivity. A n early transient component may be more sensitive to 4 A P than the slower sustained component. Consequently, L S O principal neurons may possess two classes of 4 A P sensitive outward rectifications that differ in their kinetics o f activation, and inactivation properties, such as a delay-type ( G D ) conductance, and slower onset A type ( G A ) conductance (Storm, 1990; Storm, 1987; McFarlane and Cooper, 1991; McCormick , 1991). Contributions of other outward conductances were investigated with the broad spectrum potassium channel blocker tetraethylammonium chloride ( T E A ) . These experiments were also conducted in 600 n M T T X (n=6). The resting potential was not significantly affected by the presence o f T E A at any dosage. However, significant effects of T E A were observed on membrane depolarizations at concentrations o f 5 m M or more. Both transient and steady-state depolarizations were significantly increased (Figure 5.18 A - B ) . A t 20 m M , T E A revealed a late depolarizing event during relatively small depolarizations (Figure 5.18 C). Wi th greater current pulses, a high threshold spike, or plateau event (HTS) was evoked. The H T S appeared to rise from a distinct inflection point in the depolarizing voltage trajectory (Figure 5.18 C , stars), and therefore is likely not related to the T P . Examination o f derived I -V relations indicate that the effects of T E A on depolarizations became prominent positive to approximately -40 mV. The threshold for the H T S , based on the voltage at inflection point, is approximately -35 mV, much higher than that for N i 2 + sensitive G C a T (Figure 5.18 D-E) . These results suggest that the depolarizing voltage sag following the T P is a manifestation o f the interplay of several intrinsic membrane conductances sensitive to 4 A P and T E A . These may include A- type potassium conductances and TEA-sensitive, possibly calcium-dependent, potassium conductances. In concert, potassium T. J. A D A M 100 conductances would contribute to spike rate accommodation and suppression and truncation of repetitive discharge of the L S O chopper response. 7.3.6.5 Conductances contributing to high threshold spikes The basis for the expression o f H T S s was investigated in a series o f experiments 2~^  2"T" employing 600 n M T T X , nominal elimination o f external C a , and C d application. In standard A C S F with or without 600 n M T T X , HTSs could not be evoked in L S O principal neurons. However, additional blockade of depolarization-activated potassium conductances with 20 m M T E A revealed the high threshold event. Under these conditions, the H T S was clearly distinguishable from the low threshold depolarization by a distinct inflection point (In T T X : Figure 3.20 C , stars; Figure 3.21 B) . The threshold for H T S activation was approximately -35 m V , and inactivation was slow, producing a plateau event that was maintained over the entire 200 ms stimulus period, and often 9+ exceeded it (Figure 5.19 B) . Subsequent addition of the calcium channel blocker, C d (50 u M ) in standard A C S F (2 m M C a 2 + ) reversibly eliminated the H T S (n=5). The low threshold calcium spike (LTS) contributing to the T P was not affected by 50 u M C d 2 + . Reduction o f external C a 2 + (200 u M , n=5) eliminated both the H T S and L T S (not shown). These data show that chopper neurons possess two distinct families o f calcium conductances, differentiable on the basis of cation blockade: A transient, low threshold Ay' 9 + transient species that generates a L T S contributing to the T P ( G C 3 T ) is blocked by N i , while a high threshold species producing high threshold calcium spikes (GC aHTs) is 2+ selectively blocked by C d at low concentrations (Huguenard, 1996; Umemiya and Berger, 1994; Viana, Bayliss and Berger, 1993a, b). High threshold calcium conductances likely provide the necessary calcium influx to activate calcium-activated potassium conductances contributing to spike rate accommodation in chopper neurons. T. J. A D A M 101 7,3.6.6 Outward conductances modulating chopper spike discharge The influence of 4 A P - and T E A sensitive membrane potassium conductances on the chopper response were of great interest in the consideration of their potential role in spike rate accommodation and modulation o f chopper responses by holding potential. Consequently, the following experiment was conducted to assess their contribution to chopper firing. While T E A could be applied to L S O principal neurons over the entire dosage range tested, only low concentrations o f 4 A P (50 u M ) could be applied in the absence of T T X . A t higher concentrations of 4 A P , recordings became unstable due to excessive spontaneous firing of L S O principal neurons and their inputs. In the presence of 50 u M 4 A P , spikes were broadened during chopper discharge evoked from various holding potentials (Figure 5.20 A - B , n=6). Spike rate accommodation appeared to be reduced in 50 u M 4 A P , as inter-spike intervals remained relatively constant throughout the spike train. Act ion potentials were evoked throughout the depolarizing current pulse under 4 A P , indicating that firing was less transient. A l l firing evoked under 4 A P was repetitive and sustained, regardless of holding potential or suprathreshold current step amplitude. Therefore, 4 A P sensitive currents contribute to both spike rate accommodation and discharge transience o f the L S O chopper responses in vitro. These observations suggest that A- type potassium conductances play a notable role in the suppression o f repetitive discharge during evokation o f chopper responses from hyperpolarized holding potentials. The influence o f T E A on chopper responses was complementary to that of 4 A P , affecting primarily slower outward conductances (n=6, Figure 5.21). A t 5 m M , discharge transience was eliminated (as it was under 4 A P ) , but in a manner distinct from observations during 4 A P application. Following the onset action potential, there was a long inter-spike interval at moderate stimulus intensities, followed by repetitive spike discharge (Figure 5.21 B) . Presumably, this interval is a consequence o f the remaining influence of 4AP-sensitive conductances. During repetitive firing, action potentials T. J . A D A M 102 gradually became taller and wider. This may be due to cumulative effects of calcium entry through high threshold calcium channels, and/or the blockade of slowly activating potassium conductances. A t 20 m M , sodium spikes triggered a high threshold plateau event. A n intervening transient voltage sag may be the result o f 4 A P sensitive transient potassium conductances remaining active after T E A application (Figure 5.22 D) . The broad-spectrum potassium channel blocker T E A , therefore, affects sustained outward conductances that contribute to the chopper response by providing a hyperpolarizing shunt that inhibits firing late in the response. Potassium conductances sensitive to T E A , therefore, likely play a role in spike rate accommodation and the termination of chopper discharge during depolarization. 7.3.7 Chopper-like discharge following current pre-pulses The in vivo experiments initially described demonstrated short-term adaptation in L S O principal neurons. This process affects both excitatory and inhibitory responses. Sustained binaural responses occur only when stimulus intensities are similar in both ears. Intrinsic mechanisms contributing to short-term adaptation were studied in a series of tissue slice experiments. Here, we emphasize the influence o f the prevailing membrane potential on the chopper response. The results may have implications for the integration of excitation and inhibition in L S O principal neurons. In the tissue slice preparation (n=25), current prepulses were employed to investigate the sensitivity of the chopper response to prior membrane potential polarization. We polarized the membrane with 200 ms prepulses o f variant magnitude, spanning the subthreshold voltage range from -110 to -50 mV. A constant-amplitude test pulse just exceeding threshold immediately followed the prepulse. The firing patterns of L S O principal neurons during the test pulse were then compared for different prepulse membrane potentials. T. J. A D A M 103 Voltage responses during prepulses and test pulses are depicted in Figure 5.22. The pre-hyperpolarizations evoked depolarizing voltage sags due to anomalous rectification and pre-depolarizations produced a small T P , followed by a sustained voltage sag (Figure 5.22 A , B) . The firing patterns evoked by the test pulse varied substantially as a function o f prepulse membrane potential. While the onset spike is well-locked to test pulse onset, subsequent spikes are delayed, and their timing is somewhat variable at peri-threshold intensities. The onset latency was fairly stable and slightly shorter for chopper responses evoked from more depolarized membrane potentials (Figure 5.22 A - C ) . Over a voltage range o f -110 to -50 m V , the onset spike latency changed by only 3.2 ± 0.9 ms for the same test pulse amplitude. This corresponds to a relation between onset latency and membrane potential characterized by a slope of ~ 0.05 ms per mV. Thus, the onset spike latency was not very sensitive to the pre-existing membrane potential. Other aspects o f the chopper response were dramatically affected by prepulse amplitude. The first inter-spike interval was shortest when the test response arose from pre-potentials close to rest (-50 to -55 mV) , and lengthened at more negative or positive pre-potentials. Spike count also changed with prepulse voltage. The highest spike counts were always obtained with test pulses evoked from rest, whereas fewer spikes were elicited after pre-polarization in either direction (Figure 5.22 B , Figure 5.22 E). Following the onset action potential evoked from rest, the first oscillation triggered a second spike (Figure 5.22 A , arrowheads). Following prepolarizations o f -70 to -50 mV, later spike emerged from the crests o f these oscillations as well. These oscillations may represent, then, a tendency of L S O principal neurons to produce the regular chopper output, even when firing o f most action potentials is prevented by insufficient stimulus strength. The oscillations faded away during a negative voltage sag that followed the onset spike evoked from more negative prepolarizations (Figure 5.22 A , T. J. A D A M 104 arrow). The sag may be the product o f a potassium conductance shunt, such as that described in the previous section, but not investigated further here. During pre-depolarizations, the sustained hyperpolarizing voltage sag evident during prepulses may also prevail during the test pulse, representing a conductance shunt that suppresses the sustained discharge. To recapitulate, the repetitive firing o f chopper responses is affected not only by depolarization magnitude, but also by the prevailing membrane potential. Onset responses are not much affected, in contrast to the later repetitive discharge. This response profile is reminiscent of the suppressed, late, sustained portion of the chopper response observed in the anaesthetized preparation when contralateral sound stimuli inhibited the ipsilaterally evoked discharge (e.g. our previous study in vivo, and Tsuchitani, 1988 a, b). 7.3.8 Synaptic potentials The observation that excitatory and inhibitory short-term adaptation interacted at the level of the L S O principal neurons in a complex manner prompted the following experiment in tissue slices. We sought to investigate the postsynaptic effects o f excitatory and inhibitory short-term adaptation, as imposed during electrical stimulation of L S O input fibers. Synaptic potentials were recorded in L S O principal neurons, while input fibers in the trapezoid body were stimulated with pulse trains to induce short-term adaptation in these inputs (n=17). ' Excitatory post-synaptic potentials (EPSPs) were reliably elicited during stimulation through the laterally placed electrode, while inhibitory synaptic potentials (IPSPs) were reliably evoked upon stimulation medial to the L S O . A l l neurons displaying this stimulation pattern (IE, n=17) also exhibited chopper patterns in spike discharge. Consequently, data is reported for L S O principal neurons. Synaptic depression was evaluated through the presentation o f short stimulus trains (10 pulses) followed by test pulses to the trapezoid body at various intervals, T. J. A D A M 105 analogous to the adaptation/recovery paradigm conducted in the anaesthetized animal. For all cells tested, IPSPs decreased in amplitude, to below control levels, and recovered thereafter, within approximately 100 ms (Figure 5.23). Surprisingly, EPSPs showed no such effect after repetitive stimulation o f the trapezoid body. Their amplitude remained stable for all delays from train termination (Figure 5.24). Repetitive electrical stimulation of the L S O principal neuron may only affect the inhibitory pathway, resulting i n depressed amplitudes of IPSPs. The excitatory pathway remained stable during similar manipulations. Therefore, short-term adaptation, as reflected by synaptic depression o f postsynaptic potentials, may only occur in the inhibitory input to the L S O principal neuron. Relating these findings to short-term adaptation during tonal stimulation, our collective findings in vitro suggest that excitatory short-term adaptation may be largely a function of spike rate accommodation imposed by intrinsic membrane properties of the chopper neuron. Conversely, short-term adaptation o f inhibition may result from decreased responsiveness in the contralateral L S O input, as well as from post-synaptic sources within the L S O neuron. T. J. A D A M 106 8. T H E DESCENDING EFFERENTS OF THE L A T E R A L O L I V O C O C H L E A R S Y S T E M A s an ancillary project, we sought to characterize the suprathreshold and subthreshold response properties o f the L O C efferents. These experiments parallel those examining spike discharge pattern and statistics, as well as subthreshold response properties in chopper neurons. 8.1 Action potential discharge in LOC efferents of the LSO The firing pattern o f L O C efferents appeared to represent an accumulation of excitation (depolarization) overtime, rising to the onset Of a spike train (Figure 6.01 A ) . For perithreshold stimulation, repetitive firing was initiated late in the response period, emerging from a slow depolarizing voltage ramp (Figure 6.01 A ) . Increased current amplitude led to a steeper ramp slope, and shortened the delay to repetitive firing. Only large current pulses raised the early voltage response to spike threshold, causing an early onset spike (Figure 6.01 A , star). This action potential was followed by a prolonged afterhyperpolarization, the decay of which yielded the depolarizing ramp preceding repetitive firing. The spike discharge of L O C efferents appeared to reflect a cumulative release from action potential inhibition imposed on the L O C efferent by the time course o f subthreshold events preceding discharge. 8.1.1 Basic membrane properties and subthreshold responses The subthreshold behaviour o f L O C efferents was characterized by apparently passive membrane responses in the voltage range hyperpolarized from rest (Figure 6.01 B) . Single exponentials corresponding to membrane time constants (Table 5.02, Table 5.03) described voltage trajectories at response onset as well as termination. In the voltage range depolarized from rest, small subthreshold current pulses also led to a simple T. J. A D A M 107 exponential rise and decay o f non-rectifying L O C efferent voltage trajectories. However, larger current pulses evoked responses dominated by a transient outward rectification. In these responses, the membrane was depolarized to an initial peak (Figure 6.01 B , star), before an apparent outward conductance initiated a rapid transient repolarization. Subsequently, this voltage sag subsided as a depolarizing ramp (Figure 6.01 B , arrow) to steady-state. The initial depolarization potential is apparently sculpted, at least in part, by the activation of the transient outward rectification, whereas the depolarizing ramp could be due to its inactivation. To illustrate the dual effects of this rectification, an inverted trace of the passive hyperpolarizing response obtained with matching current intensity (Figure 6.01 B , grey trace) is superimposed on the record o f the depolarizing response. Evidently, the L O C efferent response is initiated by passive membrane mechanisms, while the subsequent voltage sag is produced by the activation o f a transient outward conductance, causing a temporary depression of the voltage response, which lasts approximately 200 ms in this example. For the present sample, the duration of the voltage sag ranged between 100 and 200 ms, depending on the amplitude of depolarizations (p<0.05). The I - V relations for L O C efferents were linear for hyperpolarizations; both during onset (10 ms) and sustained (195 ms) responses (Figure 6.01 C). Slope conductances were not significantly different between the early and steady-state responses. However, for depolarizations, the transient voltage sag was evident positive to -60 m V , as a decreased slope (Figure 6.01 C , arrows, measured 10 ms after stimulus onset) when compared to the steady-state I -V function. 8.1.2 The temporal pattern and discharge statistics ofLOC efferent firing The temporal pattern of L O C efferent firing was initially examined in the P S T H and ISIH. L O C efferents exhibited flat P S T H s for repetitive firing (Figure 6.02 A ) . This T. J. A D A M 108 suggested poor locking of spike latencies to stimulus onset, and/or highly variable spike timing. However, ISIHs were narrowly distributed, indicating regular repetitive discharge (Figure 6.02 B) . This suggests that, in fact, the apparently poor timing precision of the L O C efferent was a consequence o f temporal jitter in the onset of repetitive firing, rather than overall spike timing variability. Min imal ISEH skew also indicated little or no spike rate accommodation in L O C efferent repetitive firing. The dot raster display also revealed that the flat P S T H distributions were due to the variability in onset timing of repetitive firing, rather than irregular spike discharge per se (Figure 6.02 C). Equal spacing in the dot raster supported regular action potential occurrences, but irregular occurrences o f the onset for repetitive firing introduced latency variability. This persisted when strong stimulation evoked an early onset spike. Conversely, the P S T H mode corresponding to the onset spike was narrow (Figure 6.02 A , arrows); usually 2 bins wide (0.1 ms binwidth). In ISIHs, the appearance of the onset spike produced a second wide mode (Figure 6.02 B , stars). Thus, in spite of the precise onset latency and stable inter-spike interval distribution, the variability in the delay to repetitive firing caused P S T H s to appear flat. Discharge statistics for L O C efferents were also distinct from L S O principal neurons (Figure 6.03). Inter-spike intervals and their respective standard deviations or SD;S (Figure 6.03 A , error bars) were stable during repetitive firing, and decreased as current intensity increased (Figure 6.03 A ) . A t all intensity levels, SDjS were less than 10 % o f their respective means. This indicates regular spike discharge. Conversely, the interval between the early onset spike and repetitive firing was more variable ( S D t typically exceeded 20 % o f the ISIj). During large current steps, the onset action potential occurred with a short latency (< 3 ms) and showed little timing variability (SDi < 10 %). A n increase in current strength led to a decrease in both onset spike latency and variability (Figure 6.03 B) . Plots of coefficients of variation for intervals (CVj) against interval latency indicated regular repetitive firing with a variable initiation o f firing (Figure T. J. ADAM 109 6.03 C) . L o w CViS (< 0.2) characterized the delayed repetitive firing, while the first interval exhibited a higher C V (> 0.2). Spike rate accommodation was minimal in L O C efferents (Figure 6.03 D). Spike rate (during delayed repetitive firing) increased approximately linearly with current strength at a slope of around 70 spikes/second/nA (Figure 6.03 E). T . J . A D A M 110 9. DISCUSSION Principal neurons of the L S O encode IIDs of high frequency sounds. Intensity disparities are thought to be encoded in the spike rate of chopper responses. B y investigating the behaviour of discharge rate during prolonged auditory stimulation, we have revealed several ambiguities in the chopper rate representation o f IID that preclude its utility as a code without consideration of L S O network properties. Further, by researching intrinsic membrane properties that might contribute to chopper response generation, we have discovered a more complex repertoire o f voltage-dependent conductances controlling L S O chopper expression than anticipated from previous studies (Wu and Ke l ly , 1991; Sanes, 1990). 9.1 Chopper-like responses are produced by LSO principal neurons The population of neurons residing in the L S O is heterogeneous. A s we were primarily interested in the L S O principal neurons, we attempted to restrict our examination to only those units whose electrophysiological behaviour was consistent with the binaural comparators. In our short-term adaptation studies and tissue slice experiments, we were able to localize each unit recorded. In the former, extracellular ejection of W G A - H R P marked each recording site. In the latter, recording electrodes were placed within the boundaries of the L S O under visual guidance. Additionally, a subset o f neurons were stained with neurobiotin to confirm the reliability o f their localization. In both preparations, only recordings conducted within the L S O were included in this report. Comparison of discharge patterns amongst units recorded allowed for further refinement o f the sample proposed to be principal neurons. Only those exhibiting chopper, or chopper-like discharge patterns were included in this group. For tissue slice T . J . A D A M 111 work, the employment of the same statistical characterization o f spike discharge as used previously to define the chopper response (Young, Robert, et al., 1988; Young, Shofner, et al., 1988), confirmed that intrinsically generated firing consisted of chopper-like responses. With the exception o f V C N stellate cells, the chopper pattern is unique to L S O principal neurons, below the level of the inferior colliculus (Tsuchitani, 1988a; Tsuchitani, 1988b; Tsuchitani, 1982; Tsuchitani, 1969; Tsuchitani and Johnson, 1985; Guinan, Guinan and Norris, 1972; Guinan, Norris and Guinan, 1972). The stimulus-response patterns o f chopper neurons also supported their identity as L S O principal neurons. A l l binaural units observed in the L S O (in vivo) were of the IE type, being excited by ipsilateral B F tones, and inhibited by contralateral B F tones. In tissue slices, electrical stimulation of bilateral L S O inputs also revealed an IE response pattern. Chopper neurons responded to stimulation o f the fiber tract lateral to the L S O with EPSPs, and stimulation o f the tract medial to the nucleus with IPSPs. This response feature is also relatively unique to L S O principal neurons at the level of the auditory brainstem (Cant and Casseday, 1986; Glendenning, Hutson, et al., 1985; Tolbert and Morest, 1982a; Tolbert and Morest , 1982b; Wenthold, Huie, et al., 1987; Moore and Caspary, 1983). Finally, the morphology o f our chopper neurons recorded in vitro corresponds closely to that o f L S O principal neurons reported in anatomical studies (Helfert and Schwartz, 1987a; Helfert and Schwartz, b; Helfert and Schwartz, 1986). Both possess fusiform somata approximately 25 x 15 um in size, and have similar dendritic arborization patterns. One to three principal dendrites course with sparse branching along the iso-frequency axis, to the margins of the nucleus. This confirmed the correspondence of our tissue slice chopper neurons to L S O principal cells. The similarities of electrophysiological and morphological characteristics o f our L S O neurons to those o f L S O principal cells provide compelling evidence for their direct correspondence. Our electrophysiological experiments therefore constitute a systematic T . J . A D A M 112 effort to characterize the behaviour of L S O principal neuron chopper responses during prolonged stimulation, and during manipulation of intrinsic membrane properties. 9.2 Generation of the chopper pattern in the LSO Chopper response generation in the L S O continues to be a challenge for researchers investigating IID coding at this level. Historically, the integration of many synaptic inputs over a passive dendritic arbor was thought to underlie chopper response generation. However, the rapid and precise onset o f the chopper response is not consistent with this relatively slow integration mechanism. Additionally, V C N chopper neurons have been described to produce entrained firing during repetitive electrical stimulation o f their inputs, rather than an intrinsically generated regular rhythm (Wu and Oertel, 1987; Ferragamo, Golding and Oertel, 1998). A similar ability has also been reported for some L S O chopper neurons (Wu and K e l l y , 1991). Therefore, chopper neurons are apparently able to dissociate rapid transient spike discharge (the onset spike) from regular repetitive firing, depending on stimulus conditions. Our delineation of the intrinsic membrane properties of L S O principal neurons provides possible mechanisms by which these seemingly disparate response characteristics can be reconciled. 9.2.1 Regulation of the onset spike A distinctive feature o f the L S O chopper response was the rapid and precise onset latency. In tissue slice recordings, onset spikes occurred during direct depolarizations at a latency of 1 to 3 ms, with exquisite timing precision ( C V <: 0.1). Onset spikes rode on the onset of a transient potential (TP) visible in subthreshold responses near the onset of depolarizations above -65 m V (Figure 5.05 A , B , Figure 5.07). This T P secured a rapid and stable latency for triggering the onset of the chopper response. T . J . A D A M 113 Our investigation of mechanisms underlying the T P support the involvement of several intrinsic membrane conductances. Reduction in the amplitude of the T P during T T X application (600 n M ) is consistent with a contribution from subthreshold sodium conductances iXjNas - Llinas, 1988). Low-threshold calcium conductances ( G c a i ) also contribute to this depolarization, as indicated by the reduction of the T P in N i 2 + (50 u M , Figure 5.13, and 5.15), but not in C d 2 + (50 u M - Figure 5.19). This supports a selective contribution from GC AT> distinct from the role o f high-threshold calcium conductances, as has been observed in other neurons (Viana, Bayliss and Berger, 1993; Umemiya and Berger, 1994; Christie, Eliot, et al., 1995). Finally, delayed anomalous rectification ( G H ) was found to be involved in the elicitation o f the T P due to its delayed voltage-dependent deactivation upon depolarization (Figure 5.15 B - C , E ; Pape, 1996; Nichols and Lopatin, 1997). Simultaneous blockade of all three conductance classes resulted in the elimination o f the T P (Figure 5.15). Therefore, the T P is produced by the interplay o f three primary membrane conductances - G H , G C A T , and G N a s -Other intrinsic membrane conductances also play a role in the dynamics o f the T P by shaping its decay, and restricting T P timing to the onset o f depolarizations. Anomalous rectification not only contributes to the magnitude o f the T P , but also promotes its timing at the onset o f the response. This was evident in the elicitation of TPs at widely varying, depolarization-dependent latencies upon perfusion o f the G H blocker, cesium (Figure 5.15 C , G) . Potassium conductances also play a role in T P timing. A- type potassium conductances shape the decay o f the T P (Figure 5.16), and thereby limit its peak to response onset. Additionally, TEA-sensitive potassium conductances (GK-TEA) suppress the occurrence o f the T P in later portions o f depolarizations. This was evident in voltage traces displaying delayed depolarizing humps under T E A (20 m M , Figure 5.19 B) . During hearing, converging EPSPs and IPSPs would interact with G H , GCAT> and G-Nas prior to spike generation. These membrane conductances could amplify EPSPs and decrease their rise time to spike threshold, securing a rapid and precisely timed onset T . J . A D A M 114 spike. Voltage dependent dynamics of G H and G C a T , both o f which are increased during evokation from negative holding potentials, could also preserve onset spike timing during fluctuations o f the membrane potential. Indeed, the preservation o f onset spike timing was evident in responses evoked from negative holding potential, or following current prepulses (Figure 5.07, 5.22). These conductances would provide a compensatory mechanism for the increased voltage trajectory to spike threshold, and preserve the chopper onset response in natural hearing. Clearly, L S O chopper neurons possess intrinsic membrane conductances that determine the shape and duration of a T P , resulting from neurotransmitter (glutamate) induced changes in membrane conductance. The first consequence o f this arrangement is an increased firing probability within a tightly restricted time period. The second consequence is a compensatory mechanism for the preservation of onset responses during variant stimulation histories (e.g. a precedent IPSP). 9.2.2 Regulation of repetitive firing In tissue slices, L S O principal neurons exhibited transient repetitive firing during direct depolarization. The rate, count and duration of action potential trains were directly dependent on the magnitude of depolarization. A t all stimulus levels, chopper responses possessed similar discharge statistics to L S O neurons recorded in vivo ( C V s 0.2 throughout spike trains). '" In the voltage responses o f many L S O neurons, subthreshold voltage oscillations were evident between and following action potentials. The peaks of these events coincided with expected latencies o f failed spikes (Figure 5.05 A ) . These oscillations likely support timing precision during chopper responses by promoting action potential occurrences at particular intervals. A similar role for G C a T and for B K ("big conductance", fast activating) -type G K c a has been found in hypoglossal motoneurons as occurrence of T . J . A D A M 115 depolarizing afterpotentials following each spike A H P (Viana, Bayliss and Berger, 1993a; Umemiya and Berger, 1994; Sah, 1996). During depolarizations from negative holding potentials, subthreshold voltage oscillations were obscured by a transient hyperpolarizing sag following the onset spike (Figures 5.16 and 5.22 A ) . Wi th the appearance o f this sag, a long first inter-spike interval was introduced. The second action potential occurred only after the sag had decayed enough to allow the membrane potential to cross spike threshold (Figure 5.22). The voltage dependent dynamics (increasing removal from inactivation with hyperpolarization) of the transient sag, coupled with its 4 A P sensitivity (Figures 5.17, 5.20) suggest that it is produced primarily by A- type potassium conductances ( G A ) . Probably due to the action of G A , spike trains were highly sensitive to membrane pre-potential. Deflections in the range of -5 m V were sufficient to suppress repetitive firing, leaving only the onset action potential. These voltage shifts fall wel l within the range o f IPSP amplitudes noted in v ivo (Finlayson and Caspary, 1989). Therefore, our findings are consistent with a possible role o f precedent or coincident IPSPs in the observed modulation of chopper responses. In addition, our study of short-term adaptation revealed that contralateral inhibition does not greatly influence the latency of chopper responses, although it does decrease the rate o f repetitive firing (Figures 4.04 A , D , and 4.08). Thus, negative shifts in membrane potential, as occur during IPSPs, produce discrete effects on the chopper response, disrupting repetitive firing while preserving the onset spike. Our results demonstrate that the tendency o f L S O principal neurons to exhibit chopper firing patterns is primarily a consequence o f the interplay o f resident membrane conductances, with a less prominent role for passive conduction o f synaptic inputs. In the listening animal, synaptic integration may provide a kick-start mechanism, triggering the occurrence of the T P and onset spike. Activated membrane conductances would then determine the properties of subsequent repetitive discharge, depending on the chopper neuron membrane potential and activation states of resident conductances. T . J . A D A M 116 In the listening animal, chopper neurons of the L S O possess intrinsic mechanisms allowing for the dissociation of onset responses from repetitive firing. Hyperpolarization predisposes these neurons to emphasize response onset. During ongoing binaural stimulation, resident conductances would continue to emphasize the onset o f an excitatory responses, while allowing repetitive discharge, and therefore the rate code of IID, to be compromised. 9.2.3 Spike rate accommodation Spike rate accommodation was a prominent feature o f chopper responses in the tissue slice preparation. The firing rate o f transient responses decreased to ~ 30% over a course characterized by two time constants of 3-8 ms and 32-97 ms. Mul t ip le time constants suggest the contribution of multiple mechanisms to spike rate accommodation. Our tissue slice work supports a role for several intrinsic membrane conductances in the phenomenon of spike rate accommodation and chopper response transience. The application o f 4 A P decreased spike rate accommodation in L S O chopper responses, and produced sustained repetitive firing throughout depolarizations (Figure 5.20). This suggests an involvement of G A in spike rate accommodation and discharge transience. Activation of this conductance during depolarizations evoked from negative holding potentials appears to terminate repetitive firing altogether. Perfusion o f slices with T E A indicated a prominent role for other potassium conductances in spike rate accommodation. Potassium conductances sensitive to T E A , including some calcium dependent types ( G K c a ) , suppressed repetitive firing in later portions of the chopper response (Figure 5.21 B) . The presence o f these shunt conductances, perhaps activated by high threshold calcium conductances (G C a HTs) provide support for a potential role of both calcium and potassium conductances in spike rate accommodation. T . J . A D A M 117 In other neural systems, G C a H i s has been found to interact selectively with the S K ("small conductance", slowly activating) -type G K c a (GSK, Umemiya and Berger, 1994; Viana, Bayliss and Berger, 1993a; Sah, 1996). Calcium dependent activation of G S K has, in turn, been associated with spike rate accommodation and discharge transience in hypoglossal motoneurons (Viana, Bayliss and Berger, 1993b). Considering the T E A sensitivity of G S K , and the presence of G C a H T s hi L S O chopper neurons, a critical role for these intrinsic conductances in the determination o f spike rate accommodation and the transience o f spike trains is plausible. Certainly intrinsic membrane properties in chopper neurons must contribute to short-term spike rate adaptation in the whole animal. However, other mechanisms may play a role in short-term adaptation of L S O principal neurons. Short-term adaptation may partly be due to synaptic depression at each stage o f the pathway from the inner hair cells to the superior olive, for both glutamatergic (excitatory) and glycinergic (inhibitory) synapses. Contributing mechanisms include the activation of metabotropic glutamate receptors, modulation of pre- and postsynaptic membrane conductances (such as those described above), and the regulation o f presynaptic neurotransmitter release. Each o f these factors would be fed forward to influence chopper responses, and each is likely to contribute to short-term adaptation at the level o f the L S O . Under binaural stimulation conditions, chopper responses appear to be fully developed only when interaural disparities are minimal. A t larger disparities, corresponding to lateralized sound incidence angles, the repetitive chopper discharge, and therefore the putative IID rate code, is compromised. This calls into question the validity o f a rate code for the azimuthal direction o f a sound source, at least at the single neuron level in the L S O . T . J . A D A M 118 9.2.4 Membrane rectification Previous investigations of L S O principal neurons in the tissue slice have described linear current-voltage (I-V) relations (Wu and K e l l y , 1991; Sanes, 1990). This has led to the assumption that few membrane conductances influence chopper neuron behaviour. Chopper behaviour has consequently been considered to be a manifestation o f the interaction o f synaptic influences with passive R C properties of a large dendritic arbor. We were also able to obtain quasi-linear I -V relations for chopper neurons in vitro under special conditions. However, significant membrane rectifications were evident in voltage traces for neurons of our sample (e.g. Figure 5.05 B , and 5.07). Matched rectification in the depolarized and hyperpolarized voltage ranges often resulted in quasi-linear I -V relations that belied their strong influence on chopper behaviour. Responses to hyperpolarization were dominated by rapid and delayed anomalous rectification (GKIR and G H , respectively). A s described above, G C A T , G H , G N a S and G A were activated on depolarization from various subthreshold levels. Prepulse data revealed that similar deviations from the chopper discharge pattern occur for depolarizations and hyperpolarizations o f the holding potential from rest. In both cases, repetitive firing is suppressed while onset responses are maintained (Figure 5.22). Matched rectifications emphasize response onset during polarizations from rest, and strong shunting conductances produce a subsequent relaxation of voltage responses toward rest. Discrepancies between our findings and those of previous studies likely originate in different methodologies used, or may reflect differences in species or neuron types which were carefully distinguished here. Initial recordings o f L S O neurons i n the tissue slice preparation have employed intracellular electrodes filled with potassium citrate or potassium acetate. Internal acetate has been reported to block anomalous rectification in other central neurons (Osmanovic and Shefner, 1987). This may have precluded the observation of depolarizing voltage sags in chopper neurons. T . J . A D A M 119 Intrinsic membrane conductances are responsible for membrane rectifications that contribute to the ability o f chopper neurons to rapidly react to changing sounds, by voltage dependent variations in repetitive firing. In the listening animal, inward currents supporting the T P greatly increase membrane conductance at the time o f firing onset. This action not only boosts E P S P amplitude, but also shortens the membrane time constant, and supports the short and precise onset spike latency. 9.3 Comparison with other auditory neurons in vitro Chopper neurons o f the L S O have traditionally been compared with V C N stellate neurons, and contrasted to V C N bushy cells, as well as M N T B principal neurons. In both the V C N and L S O , chopper neuron electrophysiology has been attributed with a role of passive temporal and spatial summation of PSPs to produce a rate code o f sound intensity or IID. Conversely, the electrophysiology of the other cell classes is considered to be specialized for producing a temporal code of sound features (e.g. phase-locking). The present series o f experiments reveals that chopper neuron electrophysiology shares characteristics o f both neuronal types. Our results shed new light on chopper responses generated in the L S O , which may or may not apply to V C N stellate neurons as well . 9.3.1 VCN stellate neurons There are several key attributes o f V C N chopper behaviour that are associated with summative integration. These properties pertain to response pattern, I -V relations, and the behaviour o f synaptic potentials. Several salient attributes o f V C N chopper neurons electrophysiology are shared with the L S O chopper pattern. Stellate neurons of the V C N possess fairly linear I -V relations, and respond to suprathreshold depolarization with tonic repetitive firing. Spike discharge rate is directly proportional to the magnitude o f the depolarization (Oertel, 1991; Oertel, 1985; Oertel, T. J. A D A M 120 1983; Manis and Marx , 1991). Non-inactivating potassium conductances activating between -40 and -30 m V have been revealed through T E A sensitivity, in the absence of 4-A P effects (Manis and Marx , 1991). This major source of rectification in V C N stellate neurons seems not to be manifested until voltage levels suprathreshold to action potential generation are reached. Otherwise, subthreshold I -V relations appeared linear and no further time dependent rectifications were apparent in these intracellular recordings. Existing membrane rectifications, therefore, appeared to affect chiefly the spike generation and determination o f spike discharge pattern. Stellate neurons of the V C N integrate ipsilateral excitation to yield repetitive firing. Electrical stimulation o f V C N afferent inputs yields EPSPs that exhibit synaptic depression over repeated presentations (Oertel, 1983; Oertel, 1985; Ferragamo, Golding and Oertel, 1998). Presumably, the convergence o f a large number of small inputs onto the extensive and passive dendritic arbor yields a steady depolarization of the somatic membrane. This depolarization rises to action potential threshold and produces a regular repetitive spike discharge (Arle and K i m , 1991; Hewitt and Meddis , 1993; Banks and Sachs, 1991; White, Young and Manis , 1994). Firing thereby reflects depolarization amplitude, incoming excitation levels, and therefore auditory stimulus level. Despite the similar discharge patterns of V C N stellate cells and L S O principal neurons, there may be several significant differences in the electrophysiological behaviour of these populations. Strong membrane rectifications are evident in the depolarizing and hyperpolarizing voltage ranges o f chopper neurons in the L S O . Apparently, a more extensive repertoire of intrinsic membrane conductances determines chopper response behaviour in the L S O . This repertoire includes, importantly, G H , G C A T , G N A S , G A , and G K T E A - The voltage dependencies these conductances emphasize the onset o f chopper responses and regulate repetitive discharge during evokation of responses from variant membrane potentials. Additionally, EPSPs evoked in L S O chopper neurons do not display synaptic depression, while IPSPs do. T. J. A D A M 121 In vivo recordings indicate that L S O principal neurons do share many response features with V C N stellate neurons. These include the discharge pattern and firing statistics. However, existing data from tissue slices highlight several possible differences in the nature o f information processing between the two cell classes that could transcend the limitations of differing recording techniques. 9.3.2 VCN bushy neurons, and MNTB principal cells Surprisingly, L S O chopper neurons were discovered to share interesting response properties with the phase-locking neurons of the auditory brainstem. These characteristics are generally associated with the functional role o f supporting temporal codes o f sound, a function not normally ascribed chopper neurons. The implications of these specializations for the functional role o f chopper neurons remain to be addressed. The phase-lockers exhibit a single onset spike during direct depolarizing pulses (Oertel, 1983; W u and Oertel, 1984; Manis and Marx , 1991; Zhang and Trussell, 1994; Reyes, Rubel and Spain, 1994; Schwarz and Pui l , 1998; W u and K e l l y , 1991; Forsythe and Barnes-Davies, 1993; Smith, 1995). A small T P occurs near the onset o f the voltage response. This event has been shown to confine the timing o f response onset to an early narrow time window (Zhang and Trussell, 1994; Schwarz and Pui l , 1998). Strong outward rectification in the depolarized voltage range introduces non-linearities to the I -V relation and limits spike discharge to a single action poteritial at response onset (Oertel, 1991; Oertel, 1985; Oertel, 1983; Schwarz and Pui l , 1998; Reyes, Rubel and Spain, 1994; Manis and Marx, 1991; Manis , 1990; Smith, 1995). Intrinsic membrane conductances o f phase-locking neurons have been demonstrated to underlie onset-only firing and are thought to support the temporal processing capabilities during auditory stimulation (Oertel, 1999; Oertel, 1997; Schwarz and Pui l , 1998; Manis and Marx , 1991; Smith, 1995). Membrane rectification in the hyperpolarized range is produced by anomalous rectification ( G K I R , and G H ) . The T P is T. J. A D A M 122 produced by the co-operative action of G N a s , G C a T , and G H . The decay of the T P and sustained depolarization level are determined by both 4 - A P and T E A sensitive potassium conductances, which suppress repetitive firing. Additionally, potassium conductances contributing to action potential repolarization and after-hyperpolarizations confer an ability to fire at high rates during auditory stimulation (Oertel, 1983; Reyes, Rubel and Spain, 1994; Manis and Marx , 1991; Zhang and Trussell, 1994). Together, these conductances produce rapid transient responses with quick recovery characteristics. These features support some degree of phase-locking. Synaptic potentials (EPSPs) recorded in V C N bushy cells and M N T B principal cells possess rapid rise/fall times, and are o f short duration. This is due, in part, to their interaction with depolarization activated intrinsic membrane conductances. Their interplay yields a short membrane time constant, amplifies the synaptic potential, and accelerates its time course (Oertel, 1983; Reyes, Rubel and Spain, 1994; W u and K e l l y , 1991). Such membrane mechanisms minimize temporal overlap with subsequent synaptic events. This preserves the size and shape of EPSPs during repetitive stimulation. Consequently, spike discharge o f phase-locking neurons recovers from evokation quickly, and elicits consistent responses to subsequent synaptic volleys. This ability strongly supports the temporal coding of the auditory stimulus. Chopper neurons o f the L S O exhibit a T P and strong outward rectification that become more prominent during stimulation from hyperpolarized membrane potentials. Act ion potential discharge from the same voltage levels resembles that o f the phase-lockers, with only onset spikes being elicited. Additionally, intrinsic membrane conductances delineated in the phase-lockers correspond to those we have found in L S O chopper neurons. Finally, EPSPs recorded in L S O principal neurons do not exhibit the expected degree of temporal summation or synaptic depression, a behaviour that has been demonstrated for the phase-lockers. Therefore, L S O principal neurons share several features with A V C N bushy neurons and M N T B principal cells, that are normally ascribed critical roles in temporal T. J. A D A M 123 processing of sound. The functional relevance of these specializations in L S O chopper neurons remains speculative, however, justified speculations would include a need to phase-lock with amplitude modulations and low frequencies of moving and stationary sounds. 9.4 Mechanisms underlying sensitivity to IIDs It is well understood that ipsilateral excitation and contralateral inhibition are integrated in the L S O principal neuron to create chopper output encoding IID. However, the qualitative differences between L S O chopper output and primary-like input discharge patterns have presented a conundrum for the modeling o f chopper response generation in the L S O . Summative passive integration o f primary-like discharge to create chopper behaviour is difficult to reconcile with the L S O chopper response itself. 9.4.1 The integration of excitation and inhibition in the LSO The classic integration model entails the summation of convergent excitation and inhibition over a passive dendritic arbor. This linear process yields a steady net polarization of the somatic membrane at the spike generation zone. Net depolarizations would be translated directly into repetitive firing once spike threshold is crossed. This is achieved via Hodgkin-Huxley-type action potential generation mechanisms (i.e. G^ar, G K ) . Through this process, the chopper discharge rate, and therefore IID code, is determined by the magnitude o f net depolarizations. Larger depolarizations result from greater excitation, relative to inhibition, and produce chopper responses with greater discharge rates, indicating presumably, more lateralized sound sources. While passive integration models do provide a plausible mechanism for the transformation o f primary-like discharge into a regularly spiking output, it is difficult to resolve linear PSP summation with several aspects of chopper behaviour, (i) The chopper T. J. A D A M 124 response possesses a distinctively rapid and precise onset. The time required for the spatial and temporal summation of PSPs over the extensive dendritic arbor of the chopper neuron should preclude this response feature, (ii) Contralateral inhibition bears preferential influence on later portions o f the chopper response (e.g. Figure 4.02). A s the contralateral intensity is raised, inhibition influences progressively earlier action potentials (Tsuchitani, 1988a; Tsuchitani, 1988b). Such temporal dependence of inhibition is inconsistent with linear summation, without the delayed arrival o f inhibition, (iii) Inhibition has been shown to disrupt spike timing precision as well as decrease discharge rate (Tsuchitani, 1988a). This, too, contradicts a linear integration information processing model, (iv) The timing precision of the onset action potential in chopper responses is not sensitive to contralateral inhibition, until the contralateral stimulus is of sufficient intensity to suppress the chopper pattern altogether (Figure 4.03, Figure 4.05). These observations are clearly difficult to explain with simple integration models entailing the spatial and temporal summation o f PSPs over a passive membrane. The rich repertoire of intrinsic membrane conductances contributing to chopper response generation must mediate the transduction of convergent excitation and inhibition. The voltage-dependent behaviour o f these conductances and their interplay must therefore be considered in models o f synaptic integration and chopper response generation. The T P produced by the interplay o f G H , G C A T , and G N A S is responsible for the distinctive onset properties o f the chopper response. The voltage-dependent dynamics of these conductances not only preserve the occurrence o f the onset action potential, but also limit its timing to the earliest portions of the depolarization. Blockade o f G H , in particular, compromised the restriction of the T P to the onset of responses (Figure 5.15). In the whole animal, the T P would conserve the onset portion o f the chopper response during variations in membrane potential, and counteract superimposed influences of convergent inhibition to preferentially maintain the onset of the chopper response. T. J. A D A M 125 Outward potassium conductances activated during depolarizations likely contribute to the temporal dependence of inhibitory effects. From negative baselines, G A suppresses repetitive firing following the onset spike, and compromises any systematic relation between depolarization level and discharge rate (Figure 5.22). Due to the voltage dependence of this conductance, evokation of chopper responses from more negative membrane potentials would decrease spike rate and timing precision in portions o f spike discharge. This conductance, in particular, may have a significant role in the temporal dependence o f inhibition, and also contribute to not only decreased chopper discharge rates, but also to compromised timing precision. A n interplay o f intrinsic and synaptic membrane conductances influences the chopper response. However, the determination of information integration mechanisms in L S O principal neurons remains a difficult task due to the lack of information regarding the relative timing o f excitation and inhibition at the spike trigger zone. These issues, considered below, remain be delineated experimentally to develop a rational model of information integration in the L S O . 9.4.2 The relative timing of excitation and inhibition during IID coding A significant issue for information integration in the L S O chopper neuron is the relative time of arrival o f EPSPs and IPSPs at the initial segment of the axon In the anaesthetized preparation, the temporal dependence of inhibitory effects on the chopper response (desynchronization and decreased rate) have been interpreted to reflect the delayed arrival o f inhibition relative to excitation (Tsuchitani, 1988a; Tsuchitani, 1988b). However, conspicuous mechanisms exist for the maintenance of temporal precision in ascending neural signals representing high frequency sounds. I f inhibition is delayed relative to excitation, and the L S O neuron simply summates convergent input, what is the purpose o f preserving the temporal structure of, and relation between, excitatory and inhibitory inputs? T. J. A D A M 126 Despite the thorough delineation o f adaptation supporting temporal precision of input activity, models of linear integration of excitation and inhibition in L S O chopper neurons have persisted. In rate code models, EPSPs and IPSPs overlap in time for their summation, but the precision of their relative timing has not been an issue. Presumably, the reason is that the time required for the spatial and temporal summation of PSPs would blur the fine temporal relation between their respective arrivals at the L S O soma. Obvious adaptations supporting the precisely timed effects o f PSPs must have significant functional relevance not accounted for by the rate code model. Certain specializations for the preservation of temporal information noted in the L S O IID circuit resemble features o f the M S O I T D circuit. The input neurons of the L S O have large diameter, myelinated axons maximizing conduction speed and precision. Calyceal endings characterize synaptic interactions producing large brief EPSPs in the pathways leading to the L S O . A l so , ipsilateral and contralateral inputs exhibit transient and temporally precise onsets. The role of these specializations is well interpreted in the phase-locked representation of sound by the synchronous arrival of convergent EPSPs. However, their functional relevance remains a mystery for the rate coding L S O . Principal neurons of the L S O possess a family o f intrinsic membrane conductances that resemble that o f phase-locking neurons. These properties generate an emphasis of response onset, which has now been demonstrated for L S O principal neurons as well. This emphasis is well explained by the intrinsic membrane properties of chopper neurons and does not require the assumption o f precedent EPSPs. We have seen that the chopper response is highly sensitive to the pre-existing membrane potential from which the onset o f depolarization, and an E P S P volley, would rise. While the onset is preserved, repetitive discharge is compromised by membrane voltage variations as small as ± 5 m V (Figure 5.22). Additionally, intracellular recordings of L S O chopper neurons during binaural stimulation in vivo reveal that IPSPs can arrive at the cell soma prior to EPSPs (Finlayson and Caspary, 1989). The hyperpolarizing and shunting action o f a precedent LPSP could T. J. A D A M 127 account for noted disruption o f chopper responses in vivo (Tsuchitani, 1988a, Tsuchitani, 1988b) through its interaction with intrinsic membrane conductances. We expect that excitation and inhibition arrive at the chopper cell soma more or less synchronously, and that their interaction with resident voltage dependent conductances determines the output pattern. While our results support the approximately coincident arrival of EPSPs and IPSPs, it is appreciated that the latency of arrival of each depends strongly on stimulus intensity, sound source location, and hearing thresholds. The voltage dependent intrinsic membrane conductances may endow chopper neurons adjustment mechanisms supporting plasticity of L S O sound coding. 9.5 Implications for the representation of IID in the LSO This dissertation research was motivated by the model o f IID coding in single units of the L S O . Studies centered on the representation of IID in the discharge rate o f chopper responses for each L S O principal neuron. Our findings in both the anaesthetized preparation and tissue slice have important implications for modeling the IID code, and the functional role of IID coding in the L S O . Several surprising results call for a re-evaluation o f current thought on IID coding. The fact that L S O principal neurons are sensitive to IIDs led to the supposition that the primary function of this circuit is the localization of high frequency sound. The azimuthal location of sound sources is encoded in the discharge rate o f the chopper response. Over the spectral map o f the L S O , various frequency components o f a complex stimulus could be interpreted by higher auditory centres on the basis of a common, perhaps average, discharge rate o f the L S O principal neurons. Since natural listening conditions entail the continuous processing o f auditory stimuli over time, we predicted that chopper rate would remain stable over the stimulus period. In the presence of short-term adaptation effects, a prolonged stationary stimulus could presumably be perceived as moving towards midline. Indeed, this research indicates T. J. A D A M 128 the occurrence of short-term adaptation in L S O chopper responses. Therefore, discharge rate cannot serve as a sufficient code for the azimuthal location of sound sources. Chopper output must be recombined wi th other information (e.g. frequency/intensity shifts, codes o f head/body motion) at higher centers (e.g. inferior colliculus) to remove ambiguities pertaining to short-term adaptation and variations in discharge rate wi th absolute stimulus intensity. Such network properties could provide a map o f azimuthal sound sources, tracking o f sound source movement, and selective attention to different sound sources by their unique locations along the azimuth. Our tissue slice studies indicate that L S O principal neurons possess intrinsic mechanisms that determine the shape, timing, and duration of the polarization resulting from neurotransmitter-induced (glutamate and glycine) membrane conductance changes. These mechanisms increase the probability o f an action potential occurrence within a very narrow time window, and compensate for prepolarization of the membrane to conserve temporal aspects of the onset response. Other aspects of the chopper response are compromised as a result of membrane prepolarization, including the L S O rate code of IID at the single neuron level. Previously, we have assumed that specializations of L S O inputs supporting the temporal coding of sound served to ensure the roughly coincident arrival of EPSPs and IPSPs at the spike generation zone o f the chopper neuron. However, the post-synaptic adaptations for the preservation o f temporal information in the output of chopper neurons suggest a more complex and remarkable functional role o f the L S O chopper than previously surmised. Why should L S O principal neurons possess most or all o f the specializations normally associated with the encoding of timing information? Recent electrophysiological studies conducted in vivo suggest that L S O principal neurons may detect the temporal decorrelation of binaural sound signals. The respective times of arrival for EPSPs and IPSPs depend not only on IID, but also on absolute stimulus intensity. This results in the sensitivity of chopper neurons to ITDs as well as IIDs (Joris, 1998; Joris, 1996; Joris, 1995; Finlayson and Caspary, 1991). Precedent T. J. A D A M 129 EPSPs result in the transient depolarization o f chopper neurons, whose magnitude and time course depends not only on IID, but also on absolute stimulus intensity. Transient discharge rate may depend on intensity, while the duration o f frring may rely on IID. The synchronous arrival o f excitatory and inhibitory inputs during equal intensity stimulation must be necessary for measuring rapid and ongoing interaural intensity fluctuations (Joris, 1998. Joris, 1996; Joris, 1995). For positive IIDs (excitation exceeding inhibition), alterations in discharge rate and transience could both encode IID. Intrinsic membrane conductances are certain to play a role in excitatory and inhibitory integration. Additionally, voltage dependent membrane mechanisms are, in principle, adjustable, and could support plasticity o f L S O sound coding. Fine tuning of intrinsic membrane conductances could contribute to compensation for hearing loss, affecting input PSP magnitude and timing, for example. The seemingly disparate new findings for the electrophysiological behaviour of L S O principal neurons would support a call for a re-evaluation of the rate code model for L S O sound coding. Present results do not resolve the ambiguity of the chopper pattern. They suggest that under prolonged stimulus conditions, as in natural listening, contralateral inhibition would largely degrade discharge regularity, a pre-requisite for chopping. The chopper pattern itself may therefore be unimportant, not encoding any sound attribute. The onset (transient) response, in contrast, may be a salient feature o f L S O principal neuron output. Transient onset responses could encode ongoing amplitude modulation, as suggested previously (Joris, 1998; Joris, 1995; Finlayson and Caspary, 1991). This dissertation research supports a novel proposition that L S O principal neurons may subserve functions beyond the localization of sound. Evidence presented here, and re-interpretation of previous research, lends support to a role in the conveyance of information pertaining to the temporal structure o f the complex binaural stimulus. Future electrophysiological recordings (in vivo) employing more complex stimulation paradigms wi l l likely provide illuminating findings that w i l l clarify the functional role for these remarkable auditory neurons. T. J. A D A M 130 9.6 The lateral olivocochlear efferents of the LSO 9.6.1 Summary ofprimary findings A second electrophysiological response class coexists with chopper neurons in the L S O , in vitro. The delay neuron responds to direct current injection with repetitive firing, that begins at a significant and highly variable delay from stimulus onset for low to moderate depolarizations. Repetitive firing occurs with a stable rate, reflecting rninimal rate accommodation. However, the response is poorly locked in time to stimulus onset. Wi th large current injections, an early, precisely time locked onset spike occurs prior to the delayed repetitive firing. Intrinsic membrane properties certainly contribute to the discharge behaviour of the delay neuron. In the hyperpolarized range, responses appear passive, without voltage- or time dependent rectification. In the subthreshold depolarizing range, however, a transient hyperpolarizing voltage sag precedes repetitive firing, whose onset is consequential to the sag decay. During large current pulses, the membrane voltage trajectory surpasses spike threshold prior to the sag onset. Thus, the onset action potential is superimposed on an initial small depolarizing event. 9.6.2 Identification of delay neurons as efferent lateral olivocochlear neurons O n the basis o f the distribution, electrophysiology and morphology of delay neurons, we have identified them as constituents o f the lateral olivocochlear system. Recordings o f L O C efferents conducted in the whole animal are rare, and confirmation of their projection patterns is lacking. Delay neurons were observed throughout the L S O , and are therefore intermingled with ChoNs. Approximately 14 % o f all neurons studied were delay neurons. In anatomical studies, L O C efferent projections in the rat originate with L S O neurons T . J . A D A M 131 comprising approximately 10 % of the population (Helfert, and Schwartz, 1987a; Helfert, and Schwartz, 1987b; Helfert, and Schwartz, 1986). Consequently, the ratio of delay neurons to the total L S O population corresponds reasonably well to that noted in anatomical studies. While a greater incidence of encountering the small delay neurons with a microelectrode was not expected, it is possible that these cells are more robust than other L S O neurons, and better survive manipulations during tissue slice preparation, resulting in a slightly higher proportion o f L O C cells being encountered. The in vitro spike discharge o f delay neurons strongly resembles that described in extracellular recordings of thin efferent fibers (presumptive L O C cells) in the spiral ganglion during acoustic stimulation (Brown, 1994; Robertson, and Gummer, 1985). Discharge patterns are repetitive and highly regular, but not locked in time to stimulus onset. Latencies to firing are long, and highly variable. Consequently, there is a good match between responses to direct depolarization, and acoustic stimulation. Further, a recent publication characterizing the in vitro electrophysiology and underlying membrane conductances of retrogradely labeled L O C neurons describes discharge behaviour and subthreshold characteristics that correspond exactly with the current data. Firing is repetitive and delayed, an onset spike occurs during large depolarizations, and a transient outward rectification is apparent during depolarizations, whereas no rectification is apparent during hyperpolarizations (Fujino, Koyano, and Ohmori, 1997). Finally, morphological characteristics of delay neurons mimic those described for the descending efferents of the L O C system. The somata are small (15 um), and approximately spherical. The dendritic arbors originate as two to three principal roots, branching into numerous fine processes which extend outwards along tortuous trajectories. Their orientation is related to the cell's position in the nucleus; extending perpendicularly to the longitudinal axis of the nucleus in the middle and lateral limbs, and approximately parallel to L S O curvature in the medial limb. This profile corresponds exactly to that described for descending efferents of the lateral olivocochlear system (Helfert, and Schwartz, 1987a; Helfert, and Schwartz, 1987b; Liberman, and Brown, T. J. A D A M 132 1986; Ryan, Schwartz, et al., 1987; Vetter, and Mugnaini, 1992; White, and Warr, 1983). Also , delay neurons were observed throughout the rat L S O , as are the L O C neurons of small rodents (Guinan, Warr, and Norris, 1983; Helfert, Schwartz, and Ryan, 1988; Helfert, and Schwartz, 1987a; Helfert, and Schwartz, 1987b; Helfert, and Schwartz, 1986; Ryan, Schwartz, et al., 1987; Schwarz, Schwarz, and H u , 1988; Vetter, and Mugnaini, 1992; Fujino, Koyano, and Ohmori, 1997). In contrast, L O C efferents are distributed only in each L S O hilus o f the cat and monkey (Warr, 1975; Carpenter, Chang, et al., 1987). It is therefore not surprising to encounter two electrophysiological cell types in the rat, distributed throughout the L S O . Consequently, all evidence compiled during the present experiments confirms the identity o f delay neurons as belonging to the L O C system. 9.6.3 Functional significance The transient rectification and delay to repetitive firing in L O C neurons during depolarization is apparently caused by transient outward conductances. The initial depolarization prior to the sag likely represents the rudiment o f passive depolarization; shortened and reduced in amplitude by the activation o f transient outward currents. Due to the action of these currents, the single onset spike is only triggered by the depolarizing event at current strengths sufficient to cause significant inactivation o f the later repetitive action potentials. T w o potassium currents, described in retrogradely labeled L O C neurons probably produce the transient outward rectification. Activation of a transient A- type current rapidly repolarizes the neuron, sculpting the initial depolarizing event, and combines with a slower delay-(D)-type current to produce the depolarizing voltage ramp (Fujino, Koyano, and Ohmori, 1997). While the ramp slope and delay to firing are determined by the relative strengths o f these currents, discharge rate is constant beyond threshold. T. J. A D A M 133 In a normal hearing animal, L O C neurons may typically exhibit a delayed repetitive discharge pattern with a stable firing rate during acoustic stimulation, based on recordings o f thin efferent fibers (Robertson, and Gummer, 1985; Brown, 1994). This corresponds well with in vitro recordings. The spike discharge o f L O C neurons appears to reflect a cumulative release from action potential inhibition imposed on it by the time course of subthreshold events preceding discharge. This may reflect a neuroprotective function for L O C neurons during intense acoustic stimulation, perhaps manipulating the gain control function o f inner hair cell transduction during prolonged intense exposure. A s the duration o f depolarization proceeds, the A - and D-type outward conductances inactivate (Storm, 1988), spike discharge is exhibited, and inner hair cell transduction modulated. In contrast, the onset spike observed in L O C neurons may only be elicited during very intense acoustic stimuli, i f at all. The exact functional effect o f the delay response pattern remains obscure, as does the contribution of the L O C system to auditory function. 10. CONCLUSIONS The broad scope of this dissertation contributes to the debate over the functional role of the L S O chopper response in the segregation of auditory stimuli. Experiments were designed to evaluate the modulation of chopper responses during prolonged auditory stimulation, and to delineate intrinsic membrane properties contributing to chopper response generation in the tissue slice. Results support the following conclusions: • The purported rate code o f IID generated in the L S O is compromised during prolonged stimulation, while onset responses are largely preserved. Therefore, chopper discharge rate at the single unit level can not serve as the exclusive cue for sound T. J. A D A M 134 processing in higher centers. Without re-combination with other information, chopper rate is not a sufficiently unambiguous code of IID. • Chopper-like responses are obtained in the tissue slice during direct depolarizations. Therefore, intrinsic membrane properties contribute to chopper response generation in the L S O principal neuron. These properties must mediate the integration of convergent excitation and inhibition to yield the chopper output. • Pharmacological experiments revealed that the intrinsic membrane properties of L S O chopper neurons are imparted by a repertoire of intrinsic membrane conductances delineated in other neuronal species. These include a low-threshold transient calcium conductance, subthreshold sodium conductance, transient potassium conductances, slower TEA-sensitive potassium conductances, high-threshold calcium conductances, and finally, both a rapid and a delayed anomalous rectifier. • The ensemble o f membrane properties and conductances described in the L S O chopper neuron resembles that for the classical temporal processors o f the central auditory system (e.g. V C N bushy neurons). Consequently, L S O chopper neurons share features with both linear summators and temporal processors. These postsynaptic specializations presumably support the encoding o f some temporal aspect of sound. • In vitro pre-pulse data support a notion that membrane non-linearities may dictate the modulation o f chopper responses by membrane potential. In natural listening, where auditory stimuli are encoded on an ongoing basis, intrinsic membrane conductances may support the compensation for mild hearing losses affecting PSP amplitude and timing at the level of the L S O . The adjustment of intrinsic conductance parameters (changing T. J. A D A M 135 channel number or voltage range, for example) would be one source of plasticity in the central auditory system. • Collectively, our results suggest that the chopper response, as studied thus far, may be a by-product of the stimulus conditions employed to study them (short tone pips). We propose that in more complex listening situations, the transient onset portion of the chopper response wi l l be revealed as a salient code-carrying component of the chopper response, in conjunction with the repetitive discharge. The stimulus aspects being encoded by individual L S O principal neurons remain to be delineated precisely. Future research efforts employing complex stimulation paradigms and modeling with neuronal assemblies w i l l be necessary to illuminate this exciting issue. T. J. A D A M 136 11. N O M E N C L A T U R E A B R auditory brainstem evoked response A C S F artificial cerebrospinal fluid A H P after-spike hyperpolarizing potential A N auditory nerve A N O V A analysis o f variance A V C N anteroventral cochlear nucleus B F best frequency C h o N chopper neuron D A P depolarizing afterpotential D e l N delay neuron E E excitatory-excitatory response for contralateral-ipsilateral stimulation E P S C excitatory postsynaptic current EPSP excitatory postsynaptic potential EO excitatory-nil response for contralateral-ipsilateral stimulation G A transient A-type potassium conductance G B K "big-conductance"-type calcium dependent potassium conductance GcaHTS high-threshold calcium conductance GcaT transient low-threshold calcium conductance G H hyperpolarization-activated cation conductance GjCCa calcium dependent potassium conductance G K I R inwardly rectifying potassium conductance G K - T E A TEA-sensitive potassium conductance G N 3 S subthreshold sodium conductance G N 3 T transient sodium conductance G S K "small-conductance"-type calcium dependent potassium conductance T. J. A D A M 137 H A P hyperpolarizing afterpotential H P holding potential H T S high threshold spike IC inferior colliculus ICc central nucleus of the inferior colliculus IE inhibitory-excitatory response for contralateral-ipsilateral stimulation IH hyperpolarization-activated cation current IID interaural intensity disparity IPSP inhibitory postsynaptic potential ISI inter-spike interval ISIH inter-spike interval histogram I T D interaural time disparity L O C lateral olivocochlear L S O lateral superior olive L T S low threshold spike M N T B medial nucleus o f the trapezoid body M S O medial superior olive PSP postsynaptic potential P S T H peristimulus time histogram P V C N posteroventral cochlear nucleus R M P resting membrane potential SOC superior olivary complex SPL sound pressure level S R A spike rate accommodation T E A tetra-ethylammonium T P transient potential T T X tetrodotoxin T. J. ADAM 138 VCN ventral cochlear nucleus OE nil-excitatory response for contralateral-ipsilateral stimulation 4 A P 4-aminopyridine T. J. A D A M 139 12. REFERENCES Abou-Madi , L . , Pontarotti, P., Tramu, G . , Cupo, A . and Eybalin, M . (1987) Coexistence o f putative neuroactive substances in lateral olivocochlear neurons of rat and guinea pig. Hearing Research 30,135-146. Adam, T. J. (1992) Evaluation of long-term effects o f early auditory deprivation: A n animal model of Otitis Media . M . A . Thesis completed at the University of Brit ish Columbia, Vancouver, B . C. Canada. Adam, T. J. , Schwarz, D . W. F. and Finlayson, P. G . (1997) Firing properties o f chopper and delay neurons in the lateral superior olivary nucleus of the rat. Experimental Brain Research Adams, J. C. (1979) Ascending projections to the inferior colliculus. Journal of Comparative Neurology 183, 519-538. Adams, J. C. (1983) Cytology of periolivary cells and the organization o f their projections in the cat. Journal of Comparative Neurology 215,275-289. Adams, J. C. and Mugnaini, E . (1990) Immunocytochemical evidence for inhibitory and disinhibitory circuits in the superior olive. Hearing Research 49, 281-298. Ai tk in , L . M . and Schuck, D . M . (1985) L o w frequency neurons in the lateral central nucleus of the cat inferior colliculus receive their input predominantly from the medial superior olive. Hearing Research 17, 87-93. Alger, B . E . (1984) Characteristics o f a slow hyperpolarizing synaptic potential in rat hippocampal pyramidal cells in vitro. Journal o f Neurophysiology 52, 892-894. Altschuler, R. A . , Fex, J. , Parakkal, M . H . and Eckenstein, F . (1984) Colocalization of enkaphalin-like and choline acetyltransferase-like immunoreactivities in olivocochlear neurons o f the guinea pig. Journal of Histochemistry and Cytochemistry 32, 839-843. Altschuler, R. A . , Parakkal, M . H . and Fex, J. (1983) Localization o f enkephalin-like immunoreactivity in acetylcholinesterase-positive cells in the guinea pig lateral superior olivary complex that project to the cochlea. Neuroscience 9, 621-630. Arle , J. E . and K i m , D . O. (1991) Neural modeling o f intrinsic and spike-discharge properties of cochlear nucleus neurons. Biological Cybernetics 64, 273-283. T . J . A D A M 140 Aschoff, A . , Mul ler , M . and Ott, H . (1988) Origin of cochlear efferents in some gerbil species. Experimental Brain Research 71, 252-262. Aschoff, A . and Ostwald, J. (1988) Distribution of cochlear efferents and olivo-collicular neurons in the brainstem o f rat and guinea pig: A double labeling study with fluorescent tracers. Experimental Brain Research 71,241-251. Banks, M . I. and Sachs, M . B . (1991) Regularity analysis in a compartmental model of chopper units in the anteroventral cochlear nucleus. Journal of Neurophysiology 65, 606-629. Banks, M . I. and Smith, P. H . (1992) Intracellular recordings from neurobiotin-labeled cells in brain slices of the rat medial nucleus of the trapezoid body. Journal of Comparative Neurology 12, 2819-2837. Bayliss, D . A . , Viana,F. , Bellingham, M . C. and Berger, A . J. (1994) Characteristics and postnatal development o f a hyperpolarization-activated inward current in rat hypoglossal motoneurons in vitro. Journal of Neurophysiology 71, 119-128. Bledsoe, S. C , Snead, C. R. , Helfert, R. H , Prasad, V . , Wenthold, R. J. and Altschuler, R. A . (1990) Immunocytochemical and lesion studies support the hypothesis that the projection from the medial nucleus o f the trapezoid body to the lateral superior olive is glycinergic. Brain Research 517,189-194. Boettcher, F . A . , Salvi, R. J. and Saunders, S. S. (1987) Forward masking of single neurons in the cochlear nucleus,. In: J. SykaandR. B . Masterton (Eds.), Auditory Pathway: Structure and Function, Plenum Press, N e w York, N . Y , U . S . A . , pp. Boettcher, F . A . , Salvi, R. J. and Saunders, S. S. (1990) Recovery from Short-Term Adaptation in Single Neurons in the Cochlear Nucleus. Hearing Research 48, 125-144. Borst, J. G . G . , Helmchen, F . and Sakman, B . (1995) Pre- and postsynaptic whole-cell recordings in the medial nucleus o f the trapezoid body o f the rat. Journal of Physiology 489, 825-840. Boudreau, J. C. and Tsuchitani, C. (1968) Binaural interaction in the cat superior olive S segment. Journal o f Neurophysiology 31,442-454. T. J. A D A M 141 Brawer, J. R. and Morest , D . K . (1975) Relations between auditory nerve endings and cell types in the cat's anteroventral cochlear nucleus seen with the Golgi method and Nomarski optics. Journal o f Comparative Neurology 160,491-506. Brawer, J. R. , Morest , D . K . and Kane, E . C. (1974) The neuronal architecture of the cochlear nucleus o f the cat. Journal of Comparative Neurology 155,251-300. Bregman, A . S. (1990) Auditory Scene Analysis: The Perceptual Organization o f Sound M I T Press, Cambridge, Mass. U . S . A . 773 pp. Brown, M . C. (1987a) Morphology o f labeled afferent fibers in the guinea pig cochlea. Journal of Comparative Neurology 260, 591-604. Brown, M . C. (1987b) Morphology of labeled efferent fibers in the guinea pig cochlea. Journal of Comparative Neurology 260,605-618. Brown, M . C. (1988) Brainstem branches o f olivocochlear axons in cats and rodents. Journal o f Comparative Neurology 278, 591-603. Brown, M . C. (1994) Antidromic responses of single units from the spiral ganglion. Journal o f Neurophysiology 71, 1835-47. Brunso-Brechtold, J. K . , Thompson, G . C. and Masterton, R. B . (1981) H R P study of the organization of auditory afferents ascending to central nucleus of inferior colliculus in cat. Journal o f Comparative Neurology 197, 705-722. Byer ly , L . and Hagiwara, S. (1988) Calcium channel diversity. In: A . D . Grinnell, D . Armstrong and M . B . Jackson (Eds.), Calcium and Ion Channel Modulation, Plenum Press, N e w York, N . Y . , p p . 3-18. Caird, D . and Kl inke , R. (1983) Processing of binaural stimuli by cat superior olivary complex neurons. Experimental Brain Research 52, 385-399. Calford, M . B . and Ai tk in , L . M . (1983) Ascending projections to the medial geniculate body o f the cat: Evidence for multiple, parallel auditory pathways through thalamus. Journal o f Neuroscience 3,2365-2380. Cant, N . B . (1981) The fine structure of two types o f stellate cells in the anterior division o f the anteroventral cochlear nucleus o f the cat. Neuroscience 4, 1909-1923. Cant, N . B . (1984) The fine structure o f the lateral superior olivary nucleus o f the cat. Journal o f Comparative Neurology 227,63-77. T. J. A D A M 142 Cant, N . B . and Casseday, J. H . (1986) Projections from the anteroventral cochlear nucleus to the lateral and medial superior olivary nuclei. Journal o f Comparative Neurology 247,457-476. Cant, N . B . and Morest , D . K . (1979a) The bushy cells in the anteroventral cochlear nucleus o f the cat. A study with the electron microscope. Neuroscience 4, 1925-1945. Cant, N . B . and Morest , D . K . (1979b) Organization o f the neurons in the anterior division o f the anteroventral cochlear nucleus of the cat. Light-microscopic observations. Neuroscience 4,1909-1923. Carbone, E . and Lux, H . D . (1984) A low voltage-activated calcium conductance in embryonic chick sensory neurons. Biophysics Journal 46, 413-418. Carpenter, M . B . , Chang, L . , Pereira, A . B . , Hersh, L . B . , Bruce, D . A n d W u , J . -Y. (1987) Vestibular and cochlear efferent neurons in the monkey identified by immunocytochemical methods. Brain Research 408, 275-80. Caspary, D . M . and Faingold, C. L . (1989) Non-N-Methyl-D-aspartate receptors may mediate ipsilateral excitation excitation at lateral superior olivary synapses. Brain Research 503, 83-90. Chimento, T. C. and Schreiner, C. E . (1991) Adaptation and recovery from adaptation in single fiber responses o f the cat auditory nerve. Journal o f the Acoustical Society of America 90,263-273. Christie, B . R. , Eliot , L . S., Ito, K . - L , Miyakawa, H . A n d Johnston, D . (1995) Different C a 2 + channels in soma and dendrites of hippocampal pyramidal neurons mediate spike-induced C a 2 + influx. Journal of Neurophysiology 73,2553-2557. Domnitz, R. H . and Colburn, H . S. (1977) Lateral position and interaural discrimination. Journal o f the Acoustical Society o f America 61, 1586-1598. Durlach, N . I. and Colburn, H . S. (1978) Binaural phenomena. In: E . C. Carterette and M . P. Friedman (Eds.), Handbook o f Perception, Academic Press, N e w York, N . Y , pp. 365-466. Fekete, D . M . , Rouiller, E . M . , Liberman, M . C. and Ryugo, D . K . (1984) The central projection o f intracellularly labeled auditory nerve fibers in cats. Journal of Comparative Neurology 229,432-450. T. J. A D A M 143 Feldman, M . L . and Harrison, J. M . (1969) The projection of the acoustic nerve to the ventral cochlear nucleus of the rat. A Golgi study. Journal of Comparative Neurology 137, 267-294. Felix, D . and Ehrenberger, K . (1992) The efferent modulation o f mammalian inner hair cell afferents. Hearing Research 64, 1-5. Ferragamo, M . J. , Golding, N . L . , Oertel, D . (1998) Synaptic inputs to stellate cells in the ventral cochlear nucleus. Journal o f Neurophysiology 79, 51-63. Festen, J. M . and Plomp, R. (1990) Effects o f fluctuating noise and interfering speech on the speech-reception threshold for impaired and normal hearing. Journal of the Acoustical Society of America 88,1725-1736. Fex, J. and Altschuler, R. A . (1984) Glutamic acid decarboxylase immunoreactivity of cochlear neurons in the organ o f Corti o f guinea pig and rat. Hearing Research 15, 123-131. Fex, J. and Altschuler, R. A . (1986) Neurotransmitter-related immunocytochemistry of the organ o f Corti. Hearing Research 22,249-263. Finlayson, P. G . and Caspary, D . M . (1991) Low-frequency neurons in the lateral superior olive exhibit phase-sensitive binaural inhibition. Journal of Neurophysiology 65, 598-605. Finlayson, P. G . (1995) Decreased inhibition to lateral superior olive neurons in young and aged Sprague-Dawley rats. Hearing Research 87, 84-95. Finlayson, P. G . and Adam, T. J. (1996) Excitatory and inhibitory response adaptation in the superior olivary complex affects binaural acoustic processing. Hearing Research 103,1-18. Finlayson, P. G . and Adam, T. J. (1997) Short-term adaptation of excitation and inhibition shapes binaural processing. Acta Otolaryngologica (Stockhholm) 117, 187-191. Finlayson, P. G . and Caspary, D . M . (1989) Synaptic potentials o f chinchilla lateral superior olivary neurons. Hearing Research 38,221-228. T. J. A D A M 144 Finlayson, P. G . and Caspary, D . M . (1993) Response properties in young and old Fischer-344 rat lateral superior olive neurons: A quantitative approach. Neurobiology of Aging 14, 127-139. Forsythe, I. D . and Barnes-Davies, M . (1993) The binaural auditory pathway: Membrane currents limiting multiple action potential generation in the rat medial nucleus of the trapezoid body. Proceedings o f the Royal Society of London. B . Biological Sciences 251,143-150. Fox, A . P. , N o w y c k y , M . C . and Tsien, R . W . (1987) Kinetic and pharmacological properties distinguishing three types of calcium currents in chick sensory neurons. Journal of Physiology (London) 394, 173-200. Fujino, K . , Koyano, K . and Ohmori, H . (1997) Lateral and medial olivocochlear neurons have distinct electrophysiological properties in the rat brain slice. Journal of Neurophysiology 77, 2788-2804. Gelfand, S. A . , Ross, L . and Mi l l e r , S. (1988) Sentence reception in noise from one versus two sources: Effects o f aging and hearing loss. Journal of the Acoustical Society of America 83,248-256. Glendenning, K . K . , Hutson, K . A . , Nudo, R. J. and Masterton, R. B . (1985) Acoustic chiasm. II. Anatomical basis o f binaurality in lateral superior olive o f cat. Journal o f Comparative Neurology 232,261-285. Glendenning, K . K . and Masterton, R. B . (1983) Acoustic chiasm: Efferent projections of the lateral superior olive. Journal o f Neuroscience 3,1521-1537. Goldberg, J. M . and Brown, P. B . (1968) Functional organization o f the dog superior olivary complex : A n anatomical and electrophysiological study. Journal of Neurophysiology 31,639-656. Goldberg, J. M . and Brown, P. B . (1969) Response of binaural neurons o f dog superior olivary complex to dichotic tone stimuli: Some physiological mechanisms o f sound localization. Journal o f Neurophysiology 32, 613. Guinan, J. J . , Guinan, S. S. and Norris, B . A . (1972) Single auditory units in the superior olivary complex: I Responses to sounds and classifications based on physiological properties. International Journal of Neuroscience 4, 101-120. T. J. A D A M 145 Guinan, J. J., Norris, B . E . and Guinan, S. S. (1972) Single auditory units in the superior olivary complex. II: Locations o f unit categories and tonotopic organization. International Journal of Neuroscience 4,147-166. Guinan, J. J. , Warr, W . B . and Norris, B . E . (1983) Differential olivocochlear projections from lateral versus medial zones o f the superior olivary complex. Journal of Comparative Neurology 221,358-370. Guinan, J. J. , Warr, W . B . and Norris, B . E . (1984) Topographic organization o f the olivocochlear projections from the lateral and medial zones o f the superior olivary complex. Journal of Comparative Neurology 226,21-27. Haas, H . L . , Hartter, S., Hermes, B . andHiemke, C . (1992) Bath Perfusion o f Slices with Drugs. In: H . Kettenmann and R. Grantyn (Eds.), Practical Electrophysiological Methods, Wiley-Liss , N e w York, N . Y , U . S . A . , pp. 129-132. Harrison, J. M . and Howe, M . E . (1974) Anatomy of the afferent auditory nervous system o f mammals. In: W. D . Keidel and W. D . Neff (Eds.), Handbook of Sensory Physiology, Springer, N e w York, N . Y , U . S . A . , pp. 284-336. Helfert, R. H . and Schwartz, I. R. (1986) Morphological evidence for the existence of multiple neuronal classes in the cat lateral superior olivary nucleus. Journal of Comparative Neurology 244, 533-549. Helfert, R. H . and Schwartz, I. R. (1987a) Morphologic evidence for the presence o f five cell types in the gerbil lateral superior olivary complex. Brain Research 6, 269-286. Helfert, R. H . and Schwartz, I. R. (1987b) Morphological features o f five neuronal classes in the gerbil lateral superior olive. American Journal o f Anatomy 179, 55-69. Helfert, R. H . , Schwartz, I. R. and Ryan, A . F . (1988) Ultrastructural characterization of gerbil olivocochlear neurons based on differential uptake o f 3-H-D-aspartic acid and wheat germ agglutinin-horseradish peroxidase conjugate from the cochlea. Journal of Neuroscience 8,3111-3128. Helfert, R. H , Juiz, J. M . , Bledsoe, S. C , Bonneau, J. M . , Wenthold, R. J. and Altschuler, R. A . (1992) Patterns o f glutamate, glycine, and G A B A immunolabeling in 4 synaptic terminal classes in the lateral superior olive o f the guinea pig. Journal of Comparative Neurology 323, 305-325. T. J. A D A M 146 Henkel, C. K . and Brunso-Bechtold, J. K . (1990) Dendritic morphology and development in the ferret medial superior olivary nucleus. Journal of Comparative Neurology 294, 377-388. Henning, G . B . (1974) Detectability of interaural delay in high-frequency complex waveforms. Journal of the Acoustical Society of America 55, 84-90. Hewitt, M . J. and Meddis , R. (1993) Regularity o f cochlear nucleus stellate cells: A computational modeling study. Journal of the Acoustical Society of America 93, 3390-9. Hil le , B . (1992) Ionic Channels o f Excitable Membranes (2nd. ed.) Sinauer Associates, Sunderland, Mass . U . S . A . 607 pp. Holt , G . R. , Softky, W. R., Koch , C. and Douglas, R. J. (1996) Comparison o f discharge variability in vitro and in vivo in cat visual cortex neurons. Journal of Neurophysiology 75,1806-14. Huguenard, J. R. (1996) Low-threshold calcium currents in central nervous system neurons. Annual Review of Neuroscience 58, 329-58. Irvine, D . R. F . (1986) Progess in Sensory Physiology: The Auditory Brainstem (D. Ottoson Ed.) Springer-Verlag, N e w York, N . Y , U . S . A . 279 pp. Irvine, D . R. F . (1987) Interaural intensity differences in the cat: Changes in sound pressure level at the two ears associated with azimuthal displacements in the frontal horizontal plane. Hearing Research 26,267-86. Isaacson, J. S. and Walmsley, B . (1995) Receptors underlying excitatory synaptic transmission in slices o f the rat anteroventral cochlear nucleus. Journal of Neurophysiology 73, 964-973. Isaacson, J. S. and Walmsley, B . (1996) Amplitude and timecourse o f spontaneous and evoked excitatory postsynaptic currents. Journal of Neurophysiology 76, 1566-1571. Jeffress, L . A . (1948) A place theory of sound localization. Journal o f Comparative Physiological Psychology 41, 35. Johnson, D . H . (1980) The relationship between spike rate and synchrony in responses of auditory-nerve fibers to single tones. Journal o f the Acoustical Society of America 68, 1115-1122. T. J. A D A M 147 Johnston, D , and W u , S. M - S . (1995) Foundations in Cellular Neurophysiology M I T Press, Cambridge, Mass. , U . S . A . 676 pp. Joris, P. X . (1996) Envelope coding in the lateral superior olive. II. Characteristic delays and comparison with responses in the medial superior olive. Journal of Neurophysiology 76, 2137-2156. Joris, P. X . , Carney, L . H . , Smith, P. H . A n d Y i n , T. C. T. (1994) Enhancement of neural synchronization in the anteroventral cochlear nucleus. I. Responses to tones at the characteristic frequency. Journal o f Neurophysiology 71,1037-51. Joris, P. X . , Smith, P. H . and Y i n , T. C. T. (1994) Enhancement of neural synchronization in the anteroventral cochlear nucleus. II. Responses in the tuning curve tail. Journal of Neurophysiology 71,1037-1051. Joris, P. X . and Y i n , T. C. T. (1995) Envelope coding in the lateral superior olive. I. Sensitivity to interaural time differences. Journal of Neurophysiology 73, 1043-1062. Joris, P. X . and Y i n , T. C. T. (1998) Envelope coding in the lateral superior olive. III. Comparison with afferent pathways. Journal of Neurophysiology 79,253-269. Kaltenbach, J. A . , Meleca, R. J. , Falzarano, P. R. and Myers , S. F . (1993) Forward masking properties o f neurons in the dorsal cochlear nucleus: Possible role in the process of echo suppression. Hearing Research 67, 35-44. Kiang, N . Y . S., Morest , D . K . , Godfrey, D . A . , Guinan, J. J. and Kane, E . C. (1973) Stimulus coding at caudal levels of the cat's auditory nervous system. I. Response characteristics o f single neurons. In: A . R. Mol le r (Eds.), Basic Mechanisms in Hearing, Academic Press, N e w York, N . Y , U . S . A . , pp. 455-478. Kiang, N . Y . S., Pfeiffer, R. R. , Warr, W. B . and Backus, A . S. N . (1965) Stimulus coding in the cochlear nucleus. Annals o f Otology, Rhinology, and Laryngology 74, 463-485. Ki ta , H . and Armstrong, W . (1991) A biocytin containing compound N-(2-aminoethyl)biotinamide for intracellular labeling and neuronal tracing studies: comparison with biocytin. Journal o f Neuroscience Methods 37,141-150. Klinke, R. and Galley, N . (1974) Efferent innervation of vestibular and auditory receptors. Physiological Reviews 54, 316-357. T. J. ADAM 148 Knudsen, E. I. and Konishi, M. (1978) Space and frequency are represented separately in the auditory midbrain of the owl. Journal of Neurophysiology 41, 870-884. Knudsen, E. I. and Konishi, M. (1979) Mechanisms of sound localization in the barn owl [Tyto alba). Journal of Comparative Neurology 133,13-21. Konishi, M. , Takahashi, T. T., Wagner, H., Sullivan, W. E. and Carr, C. E. (1988) Neurophysiological and anatomical substrates of sound localization in the owl. In: G. E. Edelman, W. E. Gall and W. M. Cowan (Eds.), Auditory Function: Neurobiological Bases of Hearing, John Wiley and Sons, Toronto, Ontario, Canada, pp. 721-745. Kuhn, G. F. (1977) Model for the interaural time differences in the horizontal plane. Journal of the Acoustical Society of America 62, 157-167. Lewis, B. (1983) Directional cues for auditory localization. In: B. Lewis (Eds.), Bioacoustics: A Comparative Approach, Academic Press, London, England., pp. 233-257. Liberman, M. C. (1980a) Efferent synapses in the inner hair cell area of the cat cochlea: An electron microscopic study of serial sections. Hearing Research 3,189-204. Liberman, M. C. (1980b) Morphological differences among radial afferent fibers in the cat cochlea: An electron-microscopic study of serial sections. Hearing Research 3, 45-63. Liberman, M. C. and Brown, M. C. (1986) Physiology and anatomy of single olivocochlear neurons in the cat. Hearing Research 24,17-36. Lindsey, B. G. (1975) Fine structure and distribution of axon terminals from the cochlear nucleus on neurons in the medial superior olivary nucleus of the cat. Journal of Comparative Neurology 160, 81-104. Llinas, R. R. (1988) The intrinsic electrophysiological properties of mammalian neurons: Insights into central nervous system function. Science 242,1654-64. Lu, S. M. , Schweitzer, L., Cant, N. B. and Dawbarn, D. (1987) Immunoreactivity to calcitonin gene-related peptide in the superior olivary complex and cochlea of cat and rat. Hearing Research 31,137-146. T. J. A D A M 149 Manis, P. B. (1990) Membrane properties and discharge characteristics of guinea pig dorsal cochlear nucleus neurons studied in vitro. Journal of Neuroscience 10, 2338-2351. Manis, P. B. and Marx, S. O. (1991) Outward currents in isolated ventral cochlear nucleus neurons. Journal of Neuroscience 11,2865-2880. Masterton, R. B., Granger, E. M . and Glendenning, K. K. (1994) Role of acoustic striae in hearing: Mechanism for enhancement of sound detection in cats. Hearing Research 73, 209-222. Matthews, G. G. (1986) Cellular Physiology of Nerve and Muscle. Blackwell Scientific Publications, Pallo Alto, California, U.S.A. 206 pp. McCormick, D. A. and Pape, H.-C. (1990) Properties of a hyperpolarization-activated cation current and its role in rhythmic oscillation in thalamic relay neurons. Journal of Physiology 431,291-318. McCormick, D. A. (1991) Functional properties of a slowly inactivating potassium current in guinea pig dorsal lateral geniculate neurons. Journal of Neurophysiology 66,1176-1189. McFadden, D. and Pasanen, E. G. (1976) Lateralization at high frequencies based on interaural time differences. Journal of the Acoustical Society of America 59, 634-639. McFarlane, S. and Cooper, E . (1991) Kinetics and voltage dependence of A-type currents on neonatal rat sensory neurons. Journal of Neurophysiology 66,1380-1391. Mills, A. W. (1958) On the minimal audible angle. Journal of the Acoustical Society of America 30,237-246. Moiseff, A. and Konishi, M . (1981) Neuronal and behavioral sensitivity to binaural time difference in the owl. Journal of Neuroscience 1,40-48. Moiseff, A. and Konishi, M . (1983) Binaural characteristics of units in the owl's brainstem auditory pathway: Precursors of restricted spatial receptive fields. Journal of Neuroscience 3,2553-2562. Moore, M . J. andCaspary, D. M . (1983) Strychnine blocks binaural inhibition in lateral superior olivary neurons. Journal of Neuroscience 3,237-242. T. J. A D A M 150 Morest, D . K . (1968) The collateral system o f the medial nucleus of the trapezoid body o f the cat, its neuronal architecture and relation to the olivo-cochlear bundle. Brain Research 9,288-311. Mountain, D . C. (1980) Changes in endolymphatic potential and crossed olivocochlear bundle stimulation alter cochlear mechanics. Science 210, 71-72. Muller , C . M . (1992) Extra- and intracellular recording in the slice. In: H . Kettenmann and R. Grantyn (Eds.), Practical Electrophysiological Methods, Wiley-Liss , N e w York, N . Y , U . S . A . , pp. 249-256. Nakajima, Y . (1971) Fine structure o f the medial nucleus o f the trapezoid body o f the bat with special reference to two types of synaptic endings. Journal o f Cellular Biology 50,121-134. Nichols, C. G . and Lopatin, A . N . (1997) Inward rectifier potassium channels. Annual Review of Physiology 59,171-191. Nuetzel, J. M . andHafter, E . R. (1976) Lateralization of complex waveforms: Effects of fine structure, amplitude, and duration. Journal o f the Acoustical Society of America 60,1339-1346. Oertel, D . (1983) Synaptic responses and electrical properties of cells in brain slices of the mouse anteroventral cochlear nucleus. Journal of Neuroscience 3,2043-2053. Oertel, D . (1985) Use o f brain slices in the study o f the auditory system: Spatial and temporal summation of synaptic inputs in cells of the anteroventral cochlear nucleus of the mouse. Journal o f the Acoustical Society of America 78, 328-333. Oertel, D . (1991) The role of intrinsic membrane properties in the encoding of auditory information in the cochlear nuclei. Current Opinions in Neurobiology 1,221-228. Oertel, D . (1997) Encoding of timing in the brain stem auditory nuclei of vertebrates. Neuron 19, 959-962. Oertel, D . (1999) The role of timing in the brain stem auditory nuclei o f vertebrates. Annual Review o f Physiology 61,497-519. Oldfield, S. R. and Parker, S. P. A . (1984) Acui ty o f sound localization: A topography of auditory space. I. Normal hearing conditions. Perception 13, 581-600. T. J. A D A M 151 Osen, K . K . (1970) Course and termination o f the primary afferents in the cochlear nuclei of the cat: A n experimental study. Archives o f Italian Biology 108,21-51. Osen, K . K . , Mugnaini , E . , Dahl , A . L . and Christiansen, A . H . (1984) Histochemical localization o f acetylcholinesterase in the cochlear and superior olivary nuclei. A reappraisal with emphasis on the cochlear granule cell system. Archives o f Italian Biology 122,169-212. Osen, K . K . and Roth, K . (1969) Histochemical localization of cholinesterases in the cochlear nucleus o f the cat. Wi th notes on the origin o f acetylcholinesterase-positive afferents and the superior olive. Brain Research 16, 165-185. Osmanovic, S. S. and Shefner, S. A . (1987) Anomalous rectification in rat locus coeruleus neurons. Brain Research 417,161-166. Otis, T., Zhang, S. and Trussell, L . O. (1996) Direct measurements of A M P A receptor desensitization induced by glutamatergic synaptic transmission. Journal of Neuroscience 16,7496-7504. Otis, T. S., Raman, I. M . and Trussell, L . O. (1995) A M P A receptors with high C a 2 + permeability mediate synaptic transmission in the avian auditory pathway. Journal o f Physiology 482, 309-315. Pape, H . - C . (1996) Queer current and pacemaker: The hyperpolarization-activated cation current in neurons. Annual Review of Physiology 58, 299-327. Parks, T. N . and Rubel, E . W. (1975) Organization of projections from nucleus magnocellularis to nucleus laminaris. Journal of Comparative Neurology 164, 435-448. r Pfeiffer, R. R. (1966) Classification o f response patterns o f spike discharges for units in the cochlear nucleus: Tone-burst stimulation. Experimental Brain Research 1, 220-235. Phillips, D . P., Calford, M . B . , Pettigrew, J. D . , A i tk in , L . M . , Semple, M . N . (1982) Directionality of sound pressure transformation at the cat's pinna. Hearing Research 8,13-28. Phill ips, D . P. (1992) Representation o f the two ears in the auditory cortex: A re-examination. International Journal of Neuroscience 16, 41-6. T. J. A D A M 152 Reyes, A . D . , Rubel, E . W. and Spain, W. J. (1994) Membrane properties underlying the firing o f neurons in the avian cochlear nucleus. Journal o f Neuroscience 14, 5352-5364. Rhode, W. S., Oertel, D . and Smith, P. H . (1983) Physiological response properties of cells labeled intracellularly with horseradish peroxidase in cat ventral cochlear nucleus. Journal of Comparative Neurology 213,448-463. Robertson, D . (1984) Horseradish peroxidase injection o f physiologically characterized afferent and efferent neurones in the guinea pig spiral ganglion. Hearing Research 15, 113-121. Robertson, D . and Gummer, M . (1985) Physiological and morphological characterization o f efferent neurons in the guinea pig cochlea. Hearing Research 20, 63-77. Robertson, D . and Winter, I. M . (1988) Cochlear nucleus inputs to olivocochlear neurones revealed by combined anterograde and retrograde labeling in the guinea pig. Brain Research 462,47-55. Romand, R. (1978) Survey o f intracellular recording in the cochlear nucleus o f the cat. Brain Research 148,43-65. Roth, G . L . , A i tk in , L . M . , Andersen, R. A . and Merzenich, M . M . (1978) Some features of the spatial organization of the central nucleus o f the inferior colliculus o f the cat. Journal o f Comparative Neurology 182, 661-680. Rouiller, E . M . , Cronin-Schreiber, R. , Fekete, D . M . andRyugo, D . K . (1986) The central projections o f intracellularly labeled auditory nerve fibers in cats: A n analysis of terminal morphology. Journal o f Comparative Neurology 249, 261-278. Rouiller, E . M . and Ryugo, D . K . (1984) Intracellular marking of physiologically characterized cells in the ventral cochlear nucleus of the cat. Journal of Comparative Neurology 225, 167-186. Ryan, A . F . , Keithley, E . M . , Wang, Z . - X . and Schwartz, I. R. (1990) Collaterals from lateral and medial olivocochlear efferent neurons innervate different regions of the cochlear nucleus and adjacent brainstem. Journal of Comparative Neurology 300, 572-582. Ryan, A . F . , Schwartz, I. R. , Helfert, R. H , Keithley, E . and Wang, Z . - X . (1987) Selective retrograde labeling of lateral olivocochlear neurons in the brainstem based T. J . A D A M 153 on preferential uptake of ^H-D-aspartic acid in the cochlea. Journal of Comparative Neurology 255,606-616. Ryugo, D . K . and Fekete, D . M . (1982) Morphology of primary axosomatic endings in the anteroventral cochlear nucleus of the cat: A study of the endbulbs o f Held. Journal of Comparative Neurology 210,239-257. Ryugo, D . K . and Rouiller, E . M . (1988) The central projections of intracellularly labeled auditory nerve fibers in cats: Morphometric correlations with physiological properties. Journal of Comparative Neurology 271,130-142. Ryugo, D . K . and Sento, S. (1991) Synaptic connections of the auditory nerve in cats: Relationship between endbulbs of Held and spherical bushy cells. Journal of Comparative Neurology 305, 35-48. Sah, P. (1996) Ca 2 + -act ivated K + currents in neurones: Types, physiological roles and modulation. Trends in Neurosciences 19,150-154. Sahley, T. L . , Kal ish , R. B . , Musiek, F . E . and Hoffman, D . W. (1991) Effects o f opioid be drugs on auditory evoked potentials suggest a role o f lateral olivocochlear dynorphins in auditory function. Hearing Research 55,133-142. Saint Marie , R. L . , Ostapoff, E . M . , Morest, D . K . and Wenthold, R. J. (1989) Glycine-immunoreactive projection of the cat lateral superior olive: Possible role in midbrain ear dominance. Journal of Comparative Neurology 279,382-396. Sanes, D . H . (1990) A n in vitro analysis o f sound localization mechanisms in the gerbil lateral superior olive. Journal o f Neuroscience 10, 3494-3506. Schwartz, I. R. (1977) Dendritic arrangements in the cat medial superior olivary nucleus. Neuroscience 2, 81-101. Schwartz, I. R. (1984) Axonal organization in the cat medial superior olivary nucleus. In: W. D . Nef f (Eds.), Contributions to Sensory Physiology, Academic Press, N e w York, N . Y , U . S . A . , pp. 99-129. Schwarz, D . W. F. and Pui l , E . (1998) Firing properties o f spherical bushy cells in the anteroventral cochlear nucleus o f the gerbil. Hearing Research 114,127-138. Schwarz, D . W . F . , Satoh, K . , Schwarz, I. E . , H u , K . and Fibiger, H . C. (1986) Cholinergic innervation of the rat's labyrinth. Experimental Brain Research 64,19-26. T. J. A D A M 154 Schwarz, D . W. F . , Schwarz, I. E . , H u , K . and Vincent, S. R. (1988) Retrograde transport of ( 3 H ) - G A B A by lateral olivocochlear neurons in the rat. Hearing Research 32, 97-102. Schwarz, D . W . F . , Schwarz, I. E . and H u , K . E . (1988) Transmitter neurochemistry of the efferent neuron system innervating the labyrinth. Journal of Otolaryngology 18,28-31. Schwarz, D . W . F . , Tennigkeit, F . , Adam, T. J. , Finlayson, P. G . and Pui l , E . (1997) Membrane properties that shape the auditory code in three nuclei o f the central nervous system. Journal of Otolaryngology Schweitzer, L . F. , L u , S. M . , Dawburn, D . and Cant, N . B . (1985) Calcitonin gene-related peptide in the superior olivary complex of the cat and rat: A specific label for the lateral olivocochlear system. Neuroscience Abstracts 11,1051. Searle,C. L . , Braida, L . D . , Davis, M . F . and Colburn, H . S. (1976) M o d e l for auditory localization. Journal of the Acoustical Society o f America 60,1164-1175. Sento, S. and Ryugo, D . K . (1989) Endbulbs of Held and spherical bushy cells in cats: Morphological correlates with physiological properties. Journal o f Comparative Neurology 280, 553-562. Shaw, E . A . G . (1974) Transformation o f sound pressure level from the free field to the eardrum in the horizontal plane. Journal of the Acoustical Society o f America 56, 1848-1861. Sherman-Gold, R. (ed.) (1993) The A x o n Guide for Electrophysiology and Biophysics Laboratory Techniques. A x o n Instruments Inc., Foster C i ty , C A . U . S . A . 282 pp. Shore, S. E . (1995) Recovery o f forward-masked responses in ventral cochlear nucleus neurons. Hearing Research 82,31-43. Siegel, J. H . and K i m , D . O. (1982) Efferent neural control o f cochlear mechanics? Olivocochlear bundle stimulation affects cochlear biomechanical nonlinearity. Hearing Research 6, 171 -182. Smith, P. H . (1995) Structural and functional differences distinguish principal from nonprincipal cells in guinea pig M S O slice. Journal o f Neurophysiology 73, 1653-1667. T. J . A D A M 155 Smith, P. H . , Joris, P. X . , Banks, M . I. and Y i n , T. C. T. (1989) Physiology and anatomy of principal cells in the cat M N T B . Society for Neuroscience Abstracts 15,746. Smith, P. H . , Joris, P. X . and Y i n , T. C. T. (1993) Projections of physiologically characterized spherical bushy cell axons from the cochlear nucleus of the cat: Evidence for delay lines to the medial superior olive. Journal of Comparative Neurology 331,245-260. Spain, W . J., Schwindt, P. C. and C r i l l , W. E . (1987) Anomalous rectification in neurons from cat sensorimotor cortex in vitro. Journal of Neurophysiology 57, 1555-1576. Spangler, K . M . , Cant, N . B . , Henkel, C. K . , and Farley, G . R. (1987) Descending projections from the superior olivary complex to the cochlear nucleus o f the cat. Journal o f Comparative Neurology 259,452-465. Spangler, K . M . , Warr, W . B . and Henkel, C. K . (1985) The projections o f principal cells of the medial nucleus o f the trapezoid body in the cat. Journal of Comparative Neurology 238, 249-261. Stevens, S. and Newman, E . (1936) The localization of actual sources o f sounds. American Journal o f Psychology 48, 297-306. Storm, J. F . (1987) Act ion potential repolarization and a fast afterhyperpolarization in hippocampal neurons. Journal o f Neurophysiology 385,733-759. Storm, J. F . (1988) Temporal integration by a slowly inactivating K + current in hippocampal neurons. Nature 336, 379-381. Storm, J. F . (1990) Potassium currents in hippocampal pyramidal cells. Progress in Brain Research 83,161-187. Strohmann, B . , Schwarz, D . W. F. and Pui l , E . (1994) Mode o f firing and rectifying properties o f nucleus ovoidalis neurons in the avian auditory thalamus. Journal of Neurophysiology 71, 1351-1360. Strutz, J. and Bielenberg, K . (1984) Efferent acoustic neurons within the lateral superior olivary nucleus of the guinea pig. Brain Research 299,174-177. Strutz, J. and Spatz, W. B . (1980) Superior olivary and extraolivary origin o f centrifugal innervation of the cochlea in the guinea pig. A horseradish peroxidase study. Neuroscience Letters 17,277-230. T. J. A D A M 156 Sullivan, M . E . and Konishi , M . (1984) Segregation of stimulus phase and intensity in the cochlear nuclei of the barn owl. Journal of Neuroscience 4,1786-1787. Sullivan, W. E . (1985) Classification o f response patterns in cochlear nucleus o f barn owl: Correlation with functional response properties. Journal of Neurophysiology 53, 201-216. Tennigkeit, F . , Schwarz, D . W. F. and Pui l , E . (1996) Mechanisms for signal transformation in lemniscal auditory thalamus. Journal of Neurophysiology 76, 3597-3608. Tennigkeit, F . , Schwarz, D . W. F . , Pui l , E . (1998) Postnatal development o f signal generation in auditory thalamic neurons. Developmental Brain Research 109(2), 255-263. Terhune, J. M . (1974) Sound localization abilities o f untrained humans using complex and sinusoidal sounds. Scandinavian Audiology 3,115-120. Thompson, A . M . and Thompson, G . C. (1991) Posteroventral cochlear nucleus projections to olivocochlear neurons. Journal o f Comparative Neurology 303,267-285. Tolbert, L . P. and Morest, D . K . (1982a) The neuronal architecture of the anteroventral cochlear nucleus o f the cat in the region of the cochlear nerve root: Electron microscopy. Neuroscience 7, 3053-3068. Tolbert, L . P. and Morest , D . K . (1982b) The neuronal architecture o f the anteroventral cochlear nucleus o f the cat in the region of the cochlear nerve root: Golgi and Niss l methods. Neuroscience 7, 3013-3030. Travagli, R. and Gi l l i s , R. A . (1994) Hyperpolarization-activated currents IJJ and IJQR in rat dorsal motor nucleus of the vagus neurons in vitro. Journal of Neurophysiology 71,1308-1317. Tsuchitani, C. (1969) Stimulus level of dichotically presented tones and cat superior olive S-segment cell discharge. Journal o f the Acoustical Society of America 46, 979-988. Tsuchitani, C. (1977) Functional organization of lateral cell groups o f cat superior olivary complex. Journal o f Neurophysiology 40,296-318. T. J. A D A M 157 Tsuchitani, C. (1978) Lower auditory brain stem structures of the cat. In: (Eds.), Evoked Electrical Act iv i ty in the Auditory Nervous System, pp. 373-401. Tsuchitani, C. (1982) Discharge patterns of cat lateral superior olivary units to ipsilateral tone-burst stimuli. Journal o f Neurophysiology 47,479-500. Tsuchitani, C. (1988a) The inhibition of cat lateral superior olive unit excitatory responses to binaural tone bursts. II. The sustained discharges. Journal of Neurophysiology 59, 184-211. Tsuchitani, C. (1988b) The inhibition o f cat superior olivary unit excitatory responses to binaural tone bursts: I. The transient chopper discharges. Journal of Neurophysiology 59,164-183. Tsuchitani, C. and Boudreau, J. C . (1966) Single unit analysis o f cat superior olive S segment with tonal stimuli. Journal of Neurophysiology 29, 684-697. Tsuchitani, C. and Johnson, D . (1985) The effects of ipsilateral tone-burst stimulus level on the discharge patterns o f cat superior olivary units. Journal o f the Acoustical Society o f America 77, 1484-1496. Umemiya, M . and Berger, A . J. (1994) Properties and function of low- and high- voltage-activated calcium channels in hypoglossal motoneurons. Journal o f Neuroscience 14, 5652-5660. Vetter, D . A . , Adams, J. C. and Mugnaini, E . (1991) Chemically distinct rat olivocochlear neurons. Synapse 7,21-43. Vetter, D . E . and Mugnaini, E . (1992) Distribution and dendritic features o f three groups of rat olivocochlear neurons. Anatomy and Embryology 185,1-16. Viana, F . , Bayliss, D . A . and Berger, A . J. (1993a) Calcium conductances and their role in the firing behavior o f neonatal rat hypoglossal motoneurons. Journal of Neurophysiology 69,2137-2148. Viana, F . , Bayliss, D . A . and Berger, A . J. (1993b) Mul t ip le potassium conductances and their role in action potential repolarization and repetitive firing behaviour of neonatal rat hypoglossal motoneurons. Journal of Neurophysiology 69, 2150-2163. Warr, W. B . (1966) Fiber degeneration following lesions in the anterior ventral cochlear nucleus o f the cat. Experimental Neurology 14,453-474. T. J. A D A M 158 Warr, W. B . (1969) Fiber degeneration following lesions in the posteroventral cochlear nucleus o f the cat. Experimental Neurology 23, 140-155. Warr, W. B . (1972) Fiber degeneration following lesions in the multipolar and globular cell areas in the ventral cochlear nucleus o f the cat. Brain Research 40,247-270. Warr, W. B . (1975) Olivocochlear and vestibular efferent neurons o f the feline brainstem: Their location, morphology, and number determined by retrograde axonal transport and acetylcholinesterase histochemistry. Journal of Comparative Neurology 161,159-182. Warr, W. B . (1982) Parallel ascending pathways from the cochlear nucleus: Neuroanatomical evidence of functional specialization. In: W. D . Nef f (Eds.), Contributions to Sensory Physiology, Academic Press, N e w York, N . Y , U . S . A . , pp. 1-38. Warr, W. B . and Guinan, J. J. (1979) Efferent innervation o f the organ o f Corti: T w o separate systems. Brain Research 173,152-155. Watanabe, T. and Simada, Z . (1971) Auditory temporal masking: A n electrophysiological study o f single neurons in the cat's cochlear nucleus and inferior colliculus. Japanese Journal o f Physiology 21, 537-549. Wenthold, R. J . , Huie, D . , Altschuler, R. A . and Reeks, K . A . (1987) Glycine immunoreactivity localized in the cochlear nucleus and the superior olivary complex. Neuroscience 22, 897-912. White, J. A . , Young, E . D . and Manis , P. B . (1994) The electrotonic structure o f regular-spiking neurons in the ventral cochlear nucleus may determine their response properties. Journal of Neurophysiology 71,,.1774-1786. White, J. S. and Warr, W . B . (1983) The dual origins o f the olivocochlear bundle in the albino rat. Journal of Comparative Neurology 219, 203-214. Wiegrebe, L . and Krumbholz, K . (1999) Temporal resolution and temporal masking properties o f transient stimuli: data and an auditory model. Journal o f the Acoustical Society o f America 105,2746-56. T. J. A D A M 159 Wu, S. H . and K e l l y , J. B . (1991) Physiological properties of neurons in the mouse superior olive: Membrane characteristics and postsynaptic responses studied in vitro. Journal of Neurophysiology 65, 230-246. Wu, S. H . and K e l l y , J. B . (1993) Response o f neurons in the lateral superior olive and medial nucleus o f the trapezoid body to repetitive stimulation: Intracellular and extracellular recordings from mouse brain slice. Hearing Research 68,189-201. Wu, S. H . and Oertel, D . (1984) Intracellular injection with horseradish peroxidase of physiologically characterized stellate and bushy cells in slices o f mouse anteroventral cochlear nucleus. Journal of Neuroscience 4, 1577-1588. Wu, S. H . and Oertel, D . (1987) Maturation o f synapses and electrical properties o f cells in the cochlear nuclei. Hearing Research 30,99-110. Y i n , T. C. T. and Chan, J. C. K . (1988) Neural mechanisms underlying interaural time sensitivity to tones and noise. In: G . E . Edelman, W. E . Gal l and W . M . Cowan (Eds.), Auditory Function: Neurobiological Bases o f Hearing, John Wiley and Sons, Toronto, Ontario, Canada, pp. 385-430. Y i n , T. C. T. and Chan, J. C. K . (1990) Interaural time sensitivity in medial superior olive of cat. Journal o f Neurophysiology 64,465-488. Young, E . D . , Robert, J. M . and Shofner, W. P. (1988a) Regularity and latency o f units in ventral cochlear nucleus: Implications for unit classification and generation of response properties. Journal o f Neurophysiology 60, 1-29. Young, E . D . , Shofner, W. P., White, J. A . , Robert, J . - M . and Voight, H . F . (1988b) Response properties of cochlear nucleus neurons in relationship to physiological mechanisms. In: G . M . Edelman, W . E . Gal l and W. M . Cowan (Eds.), Auditory Function: Neurobiological Bases o f Hearing, John Wiley and Sons, N e w York, N Y . U . S . A . , pp. 277-312. Zhang, S. and Trussell, L . O. (1994) A characterization o f postsynaptic potentials in the avian nucleus magnocellularis. Journal o f Neurophysiology 72, 705-718. T. J. A D A M 160 Table 5.01 Ion equilibrium potentials for whole cell patch recordings at 22 °C. Ion (RT)/F Z [ion]out [ionji,, Eton Na2+ 25.438 +1 151.25 10.3 68.36 Ca 2 + 25.438 +2 2.00 10"8 243.15 25.438 + 1 2.50 135 -101.49 cr 25.438 -1 135.50 22 -47.54 Nemst equation: E I o n = ( 2 . 3 0 3 ) ^ l o g 1 0 f[ion] ( [ion] in / where R is the gas constant, T is temperature (°K), Z is the ion's valence, and F is Faraday's constant. Ion equilibrium potentials were estimated using the Nernst equation for all free ions in solution during whole cell patch recordings conducted at 22 °C, and 100 % dissociation. Free internal calcium concentration was estimated to be 10 n M when chelated by 10 m M E G T A (see Methods). T. J. A D A M 161 Table 5.02 Basic electrophysiological properties distinguishing chopper and delay neurons in sharp microelectrodes conducted at 34 °C. C H O P P E R N E U R O N D E L A Y N E U R O N Number o f Cells 76 34 B A S I C P R O P E R T I E S Resting potential (mV) -61.21 ± 4 . 8 1 -61.67 ± 5 . 4 2 n.s. Onset time constant (ms) 2.32 ± 0 . 9 3 5.50 ± 3 . 2 1 p<.05 Input resistance (MQ) 28.14 ± 8.03 * 60.81 ± 16.77 p<01 Spike amplitude (mV) 64.85 ± 6 . 1 5 70.91 ± 7.98 p<.01 Spike half-width (ms) 0.64 ± 0 . 1 8 0.72 ± 0.22 n.s. The final column gives p-values for significant differences between the cell classes; n.s. denotes the absence of a significant difference. The symbol * denotes a significant difference between parameters for whole cell and intracellular recordings (p<0.01). T. J. A D A M 162 Table 5.03 Basic electrophysiological properties distinguishing chopper and delay neurons in whole cell patch recordings conducted at 22 °C. C H O P P E R N E U R O N D E L A Y N E U R O N Number o f Cells 102 24 B A S I C P R O P E R T I E S Resting potential (mV) -61.95 ± 5 . 3 0 -66.64 ± 8.61 n.s. Hyperpolarizing onset tau (ms) 8.44 ± 4.25 23.47 ± 7 . 5 4 p<.05 Depolarizing onset tau (ms) 8.46 ± 4 . 8 0 21.86 ± 8 . 5 7 p<.05 Input resistance ( M Q ) 109.05 ± 6 4 . 3 8 * 308.99 ± 7 1 . 1 8 p<.01 Spike amplitude [RMP] (mV) 81.27 ± 15.31 81.84 ± 16.94 n.s. Spike amplitude [inflection] (mV) 51.48 ± 16.00 45.61 ± 12.99 n.s. Spike threshold (mV) -30.14 ± 7 . 7 7 -19.49 ± 6 . 5 8 p<01 Spike half-width (ms) 0.95 ± 0 . 5 8 1.26 ± 0 . 7 6 n.s. The final column gives p-values for significant differences between the cell classes, n.s. denotes the absence o f a significant difference. The symbol * denotes a significant difference between parameters for intracellular and whole cell recordings (p<0.01). T. J. A D A M 163 Figure 3.01 The auditory brainstem in transverse section Both the M S O and L S O , as well as their primary inputs, are arranged in approximately transverse section. From the ipsilateral ear, a glutamatergic projection originating in the spherical bushy cells of the A V C N (dark grey) provides excitatory innervation o f L S O neurons. Inhibition arrives via a disynaptic pathway originating with the globular bushy cells of the contralateral V C N (dark grey). This projection terminates in the M N T B on principal neurons which then provide direct glycinergic inhibition of L S O principal neurons (light grey). T . J . A D A M 1 6 4 T. J. A D A M 165 Figure 4.01 Horseradish marks within the superior olivary complex Horseradish peroxidase marks for the track of a single electrode penetration are evident in a single histological transverse section (50 mm) through the auditory brainstem. Three extracellular recording sites were located along this track. Darker areas (increased marker diffusion) denote recording sites. One recording session was conducted in the trapezoid body ventral to the L S O (filled arrowhead), and two sessions were from units in the dorsal flexure o f the L S O (filled arrow, S-shaped nucleus). The medial nucleus o f the trapezoid body is indicated by an open arrow. Calibration bar 500 mm T. J. A D A M 166 T. J . A D A M 167 Figure 4.02 Effects of contralateral inhibition on ipsilateral excitation The chopper response is evoked during ipsilateral excitation, and is suppressed by contralateral inhibition. A . Increasing ipsilateral intensity produced greater chopper discharge evident in PSTHs . B. Increasing contralateral intensities decreased chopper discharge. C . The onset latency o f chopper firing decreased monotonically with ipsilateral stimulus intensity. However, the range of latencies remained narrow (< 4 ms). D. The onset latency of chopper responses was unaffected by inhibition. The latency remained at 4.5 ms for this neuron, over increasing contralateral intensities. T. J. A D A M 168 A Ipsilateral stimulus only 39 dB S P L riT^T , ' l H'frl o 50 '100 3 4 d B S P L 30-f 20-1 i I M M rv\ o-50 l100 29 dB SPL 0 j^ifciL 0""" n ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ | H f l " 100 Time (ms) B Ipsilateral and contralateral stimuli C/3 30-<L> 13 •c --<-> u. 4> o m 20-s? sch 10-Q pike pike CO o -Ipsi39dBSPL 30 20 H io H l50 'l00 Ipsi39dBSPL Con 24 dB SPL 30n 20 H H 0 50 100 Ipsi39dBSPL Con 31 dB SPL 1 \ ^ ^ ^ m ^ \ \ \ A \ 0 '50 'l00 C 10-i 8 -C .3 4-5'-0-1 Time (ms) Spike discharge onset latencies 10 -i Ipsilateral only 8-6-4-2-—•1 1 T -20 25 30 35 ~f f 1 40 45 50 Ipsilateral and contralateral °-tr r 00 05 10 T -15 ~T~ 20 'T-IS ~r 30 ~1 35 Ipsilateral intensity (dB S P L ) Contralateral intensity (dB S P L ) T. J . A D A M 169 Figure 4.03 Short-term adaptation in ipsilateral excitation of a binaural LSO unit Short-term adaptation in excitation is dependent on the intensity o f the ipsilateral stimulus. A . Peri-stimulus time histograms for responses to monaural ipsilateral tone presentations of varied intensity (29 to 44 dB SPL, 16.62 k H z , 50 ms duration) illustrate that both the onset peaks {filled arrows) and the steady-state response levels (*) increased with ipsilateral intensity. Additionally, the ratio o f onset to sustained responses increased, reflecting greater short-term adaptation in spike discharge. B . The magnitude of short-term adaptation in excitation increased with ipsilateral stimulus intensity. C. The time course of excitatory short-term adaptation became more rapid for stimulation at higher levels, reflecting a greater onset response for more intense stimuli. D . There was a systematic decrease in the onset latency of chopper responses with intensity increments up to 25 dB SPL from response threshold. However, the overall decrease was under 3 ms. T. J. A D A M 170 A 44 dB SPL 11 f 1111111 '50 '100 50 l100 34 dB SPL 50 '100 30 H 20-29 dB SPL 10- ^ , 0 ~"~ n " ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ 1 50 '100 Time (ms) B "u100 - l c o »-* o -a 80 6 0 H ^ 4 0 H 1 2 0 H 8P S o J Magnitude i — r T 20 25 30 35 40 45 50 Ipsilateral intensity (dB S P L ) 50 4 0 1 3 0 H | 20 10 0 4 T i m e course 1 1 1 1 1 1 1 20 25 30 35 40 45 50 Ipsilateral intensity (dB S P L ) D I O - I | 8 g 6 c co a 2 0 4 Onset latency I 1 1 1 1 1 1 20 25 30 35 40 45 50 Ipsilateral intensity (dB S P L ) T. J. A D A M 171 Figure 4.04 Short-term adaptation in contralateral inhibition of a binaural unit Short-term adaptation of inhibition in binaural L S O neurons is dependent on the intensity of the contralateral stimulus. A. Br ief ipsilateral tones (BF, 30 ms, 37 dB SPL) presented at various delays from the onset of a contralateral adapting tone (BF , 200 ms, 17 dB SPL) evoked progressively greater spike discharge (spikes/bin) for longer delays. This indicated short-term adaptation in contralateral inhibition. B. The magnitude of inhibitory short-term adaptation increased with contralateral stimulus level, while the time constant of short-term adaptation shortened from 9.52 to 7.26 ms (Con 12 to 17 dB SPL). C. Excitatory short-term adaptation during ipsilateral stimulation is depicted for the same unit as in B (17 dB SPL). The time course o f excitatory short-term adaptation (x A = 6.67 ms) was similar to that for inhibition (xl7dB = 7.56 ms), when presented at the same stimulus level (17 dB SPL). T. J. A D A M 172 Inhibitory adaptation 1 2 1 l m s I p s i 3 7 d B S P L 8 - Con 17 dB S P L o "3 4 0 II I H I HI I I I ! 2 ms <™ 1 12 -n 8 41 0-4 i — r 4 ms co "c3 ' £ 12 O m 8 - | | «-. 83 o-ex ^ 1 2 - , & 4 - 1 0 T — r 8 ms mi i TI—I 1 T 16 ms inn immJ T T C O cx 50 | u . • O 8 - | £? 4 - | •S o-t co 1 J J U 32 ms 1 1 2 - j 8 -4 -0 Li i — r 64 ms mi i II • 12. 8-4 -0-j J128 ms _ B I IB I l l l l I I BU i — i — r j i — r 0 50 100 1150200ms Time (ms) B Inhibitory adaptation time course •§160-1 U 1 4 0 H —^ • § 1 2 0 H siooH CX co a o CX C O <U Pi 80 H © Ipsi 37 dB SPL Con 17 dB SPL • Ipsi 37 dB SPL Con 12 dB SPL 12- i C ^ i o -M CX co P ? o co 8H 6 4 2H 0 50 100 150 Delay to ipsi probe (ms) Excitatory adaptation time course Ipsi 17 dB SPL 200 I 0 50 T 100 Time (ms) T 150 200 D 1 0 - , 6 C CU co O 2-1 OA Onset latency Ipsi 17 dB S P L 1 1 1 1 1 T 0 50 100 150 200 250 Probe delay (ms) T. J. A D A M 173 Figure 4.05 Short-term adaptation in binaural responses of a typical LSO unit Short-term adaptation in L S O unit responses to binaural stimulation is evident only when IIDs are positive to zero. A . During binaural stimulation at IIDs of zero, P S T H s were flat. Discharge rate (spikes/bin) was stable throughout binaural stimuli, regardless o f the overall intensity {bottom PSTH to top). B . Unequal IIDs were produced by increasing the ipsilateral stimulus intensity while holding the contralateral intensity constant. A s IID was raised {bottom PSTH to top), an early peak appeared in the P S T H , and discharge rate (spikes/bin) during binaural stimulation increased. Short-term adaptation was evident. T. J. A D A M 174 Equal interaural intensities B Unequal interaural intensities 45 dB S P L 45 dB S P L 36 dB S P L 36 dB S P L 50 100 12 " 10 -•3 8" cl co f. -o go 4-u 2 -CO 5 0-Ipsi Con 150 17 dB S P L 17 dB S P L 50 100 Time (ms) 150 200 200 200 0 12 10 8H 6 4-2-0 0 12 " 10 -8~ 6 4 2 0 r o Ipsi 45 dB SPL Con 17 dB SPL 100 150 200 Ipsi 36 dB S P L Con 17 dB S P L LllHlllhilllllllJIIlllll |j| 50 100 Ipsi Con 150 17 dB S P L 17 dB S P L 200 !' ' " " i i mm \ i "in II in 50 100 Time (ms) 150 T 200 T. J. A D A M 175 Figure 4.06 Recovery from short-term adaptation for equal interaural intensities Due to similar magnitudes and time courses of excitatory and inhibitory short-term adaptation in binaural units, responses to binaural stimulation are stable under equal-intensity (IID = 0) conditions. A . Recovery for short-term adaptation in excitation. A P S T H obtained during the first 30 ms of a 200 ms ipsilateral adapting tone shows short-term adaptation in excitation (left column). The response during a 30 ms ipsilateral probe tone presented 1 ms following the adapting tone (total probe delay = 201 ms, middle column) was still decreased. Within 256 ms (total probe delay = 456 ms, right column), responses were fully recovered, as indicated by discharge rates (spikes/bin) equal to those during the first 30 ms o f the response to the adapting tone. B . Recovery from short-term adaptation during binaural stimulation with equal interaural intensities. Short-term adaptation was not evident in responses to binaural adapting tones, and responses were minimal. N o changes in response level were evident following the adapting tone. Consequently, binaural responses were preserved during binaural stimulation where the IID was zero. C. Recovery from short-term adaptation in inhibition. Min ima l spike discharge was recorded in the P S T H during the presentation o f a contralateral adapting tone. Relative to responses to binaural adapting tones o f the same intensity (C, left column), responses during a probe tone presented 1 ms after the adapting tone were increased. Within 256 ms, responses to probe tones were restored to their original levels. D . Recovery time courses for excitatory, inhibitory and binaural short-term adaptation were similar for equal IIDs. In this unit, recovery from ipsilateral adaptation had a time constant of 62.7 ms. Inhibitory responses recovered along a course with the time constant of 51.1 ms. Resultant recovery functions are symmetrical. Consequently, responses to binaural stimulation were stable at a low discharge rate (spikes/bin). T. J. A D A M 176 A 200 ms adapting tone 1 2 Ipsi 17 dB SPL CO 'B © r o CO <4 M *&, CO L u i J 12-! 8 4H 0-T 1 1 1 20 40 Ipsi 17 dB SPL Con 17 dB SPL «j si o CO p 12-i 8 4 H 0' I r 20 ~[ ' 40 Con 17 dB SPL 1 " 1 1 1 1 0 20 40 Probe delay = 201 ms 12-i 8 4 H 0-Ipsi 17 dB SPL 200 1 T 220 240 12-i 8 4 H 0-Ipsi 17 dB SPL Con 17 dB SPL 200 220 240 12-i Ipsi 17 dB SPL i Con 17 dB SPL 4 l — r -200 220 "1 1 240 Time from adapting tone onset (ms) Probe delay = 456 ms n3 4 -0 -Ipsi 17 dB SPL r 460 T 480 500 12 1 8 -4 -0' Ipsi 17 dB SPL Con 17dBSPL kJk, -\—r 460 480 ">—r 500 12 3 0-Ipsi 17dBSPL Con 17 dB SPL M i 460 480 500 D 8-, .a • CO t 6 co ^—^ 4H too & xi o CO • 1—H T 3 » 2 2 5 1 1 1—I I I I 3 4 5 6 7 8 9 "T 1 1 1 I I I I 3 4 5 6 7 8 9 1 0 1 0 0 Delay between adapting and probe tones (ms) - i 1—r 3 4 5 T. J. A D A M 177 Figure 4.07 Recovery from adaptation for unequal interaural intensities Due to dissimilar magnitudes and time courses of excitatory and inhibitory short-term adaptation, responses to binaural stimulation with unequal positive IIDs exhibit short-term adaptation. Recovery functions for two typical L S O units are displayed. A . For moderate IIDs (< 10 dB SPL), the time course o f recovery for excitation and inhibition were not symmetrical, reflecting intensity-dependent adaptation effects. Recovery was more prominent for the more intense (ipsilateral) L S O input due to previous greater short-term adaptation. Binaural responses exhibited recovery functions resembling those for the more intense excitatory input. B . Similar results were obtained during stimulation with larger IIDs (> 10 dB SPL). However, the time course o f binaural recovery was not related to either monaural recovery function. Decreased binaural responses recovered along widely varying time courses that could not be predicted from recovery functions of excitation or inhibition. T. J. A D A M 178 ; | 100-C/3 C/3 Moderate disparities favouring the ipsilateral ear Ipsi 40 dB S P L Con 35 dB S P L 40' 20 H 0-1 • Excitation A Inhibition M Binaural 10 100 Delay between adapting and probe tones (ms) Large disparities favouring the ipsilateral ear Delay between adapting and probe tones (ms) T. J. A D A M 179 Figure 4.08 Onset latencies for binaural responses during recovery from adaptation The onset latencies of chopper responses vary only with the adaptation state of the ipsilateral excitatory input. For all IIDs, chopper onset latencies decreased during recovery of the ipsilateral input from short-term adaptation. Inhibition had no affect on chopper latencies at any contralateral intensity. T. J. A D A M 180 Delay between adapting and probe tones (ms) T. J. A D A M 181 Figure 4.09 Short-term adaptation and recovery of excitation in a monaural LSO unit Monaurally driven units of the S O C exhibit short-term adaptation and recovery. A . Discharge rate decreased rapidly during an ipsilateral excitatory adapting tone (spikes/bin, PSTH on left). Responses were decremented by 64 % along a course with time constant 5.15 ms. Responses to the probe tone presented 1 ms after the adapting stimulus remained decreased (centre PSTH), and recovered to original levels within 256 ms (PSTH on right). B. Exponential functions for the recovery from short-term adaptation in monaural units are more rapid than for binaural L S O units. For this unit, the time constants for recovery were 43.0 ms (30 dB SPL) and 15.7 (40 dB SPL). T. J. A D A M 182 A 200 ms adapting tone Delay 201 ms Delay 456 ms 0 50 100 150 200ms 200 250ms 450 500ms Time (ms) Delay between adapting and probe tones (ms) T. J. A D A M 183 Figure 5.01 The recording chamber employed for in vitro recordings The recording chamber setup employed for in vitro experiments maximized the permeation of the slice, and allows for bath temperature control. In the recording chamber, the tissue slice was suspended midstream on a nylon mesh screen, and fixed in place by a nylon thread and platinum "harp". The artificial cerebrospinal fluid entered one end of the chamber via a length of tygon tubing coiled up in a heated water bath, which warmed the medium prior to entry. The A C S F was removed via aspiration from a separate, but connected, reservoir at the distal end o f the recording chamber. Bath temperature was monitored in the recording chamber, using a thermistor adjacent to the slice mount. T. J. A D A M 184 T. J. A D A M 185 Figure 5.02 Morphology and localization of chopper and delay neurons in the LSO Camera lucida drawings illustrate the morphology and localization of representative neurons from the two electrophysiological cell types recorded in the L S O in vitro. The outline of the nucleus margins is also shown. A . Chopper neurons possessed large fusiform somata, typically located near the midline of the nucleus' axis. Dendritic arbors were bipolar, and oriented perpendicular to the axis of the nucleus curvature. T w o to three smooth principal dendrites emerged from the soma and branch sparsely, forming processes that reached across the width o f the nucleus (parallel to the isofrequency bands). B. Delay neurons had smaller spherical somata that were distributed across the nucleus band. The distribution of dendrites tended to be bipolar, with two or three principle limbs branching several times into fine processes with a tortuous appearance. The orientation of the dendritic arbor in the medial limb o f the L S O differed from the rest of the nucleus, lying oblique to the nucleus curvature. Scale bar = 100 /um. T. J. A D A M 187 Figure 5.03 Temporal pattern of spike discharge for chopper neurons Peristimulus time histograms and interspike interval histograms display the temporal discharge pattern o f the chopper neuron in Figure 5.06. Histograms were constructed from the neuron's voltage responses to 50 presentations of a 200 ms suprathreshold depolarizing current pulse. Three stimulus amplitudes are depicted. A. Peristimulus time histograms were multi-modal, wi th the first peak being the most narrow, for all stimulus levels. Spike timing variability increased slightly as the spike train proceeded, as indicated by widening of the modes. B. A dot raster display corresponding to the peristimulus histogram above shows spike rate accommodation and the associated gradual reduction of discharge timing precision during the chopper response. C. Interspike interval histograms were uni-modal. Intervals occurring after 25 ms {grey bars) produced a positive skew in the interval distribution, indicating spike rate accommodation. T. J. A D A M 188 ^ 5 0 co i4(H +-< o £ 3 0 1 § 2 0 o A. P S T H s 0.5 nA 0 [ 1 ' ' ' I 50l CO « 401 ° , 30i o o 20i 50 0.7 nA F | I f I I J I' I T'l p J I I | I | ! T| 100 150 200 Spike latency (ms) 50 1 401 8 30 I 201 o o •S 101 CQ 0 B I 50 0.9 nA I'' | I' I I | I I I I | I I I I | I I I ! | 100 150 200 Spike latency (ms) c. 201 0.5 nA ISIHs c 3 o i o 4 g l 0 201 Cj n 1 1 1 1 1 1 1 1 50 0.7 nA 11 i 11 i I 11 i 11 100 Interval (ms) 0 201 d C 50 i 1 1 1 1 1 1 1 1 • i " 1 1 100 150 200 Spike latency (ms) i 11 i i 11 i 11 50 0.9 nA 111 1 1 1 ' i 100 Interval (ms) J J ! L I I I | I I I I | I I I I | I I I I j i i i i | 0 50 100 x Interval (ms) T. J. A D A M 189 Figure 5.04 Spike discharge statistics for chopper neurons Chopper neuron spike discharge statistics reflect superior spike timing stability at all direct current stimulus intensities. A. Interspike intervals and associated timing variability (error bar = standard deviation) increased over the discharge period. Inset: Onset spike latency and timing variability (error bar = standard deviation) shortened with current amplitude increments. B. The coefficient of variation increased only slightly as a function o f interval latency, remaining below 0.2. C. Instantaneous spike rate (normalized to the first interspike interval in the train) decreased approximately 30 % along a double exponential time course (xj = 3-4 ms, x 2 = 33-49 ms). D. Average instantaneous discharge rate (over the initial 25 ms o f the response) and spike count increased monotonically with current pulse amplitude. T. J. A D A M 190 20-15-•310-fc 5H o-l Interval-Latency Relation a I 2 3 1 Onset Spike Latency i 1 1 1 i 1 1 1 i 0.6 0.8 1.0 1.2 1.4 Current (nA) 1.5 nA 1.3 nA 1.1 nA B o 1.0 - i - 1 — . — [ -50 M e a n latency (ms) go.8 H •*•> •* S3 0.6 H > | 0.4 -l-H *g0.2 H o O 0.0 H Discharge Regularity 1.5 nA 1.3 nA 1.1 nA 1 ' 100 C £ 1 2 0 -g 100-U of I 80H I 60-TJ 40-<D H ^ 20-D r o ;-.300 3 I I r -50 100 150 M e a n latency (ms) — i — i — | — i — i — i — 150 Rate Accommodation 200 200 l — i — i — i — | — i — i — i — r — | — i — i — i — i — | — r 0 50 100 150 M e a n latency (ms) - T " T - i 200 250 H 5? 200 150 H | 100 I 50 •oo I Input- Ouput Functions 02 u CD T3 ro co o o Rate Count r 30 25 r 20 7 1 5 7 10 - 5 , , ! ! r 0.6 0.8 1.0 1.2 1.4 Current (nA) 1.6 1.8 f- 0 2.0 T. J. A D A M 191 Figure 5.05 Intrinsic properties of chopper neurons in the LSO The majority o f neurons recorded in the L S O exhibit firing similar to the chopper response, as defined in vivo. This is depicted for one chopper neuron. A . Spike discharge was transient and repetitive, with discharge duration and rate increasing with stimulus intensity. Damped voltage oscillations were visible where spike threshold was not attained (arrows). The onset action potential occurred at the peak of a transient depolarizing potential, as illustrated by responses obtained during perithreshold current injections of identical amplitude (inset). B. Subthreshold voltage responses were characterized by well-matched membrane nonlinearities for depolarizations and hyperpolarizations. A transient depolarizing potential was evident at the onset o f the voltage response (star), which was o f similar amplitude and time course to peak hyperpolarizing responses. Dashed lines indicate time periods employed to construct current-voltage relations. C . Despite rectification apparent in the peak-sag profile of voltage responses, current-voltage relations were nearly linear for onset (unfdled circles) and steady-state (filled circles) responses. However, steady state responses possessed decreased slope resistances over peak current-voltage relations. T. J. A D A M 192 T. J. A D A M 193 Figure 5.06 Chopper spike discharge shows voltage dependence The chopper discharge pattern is influenced by membrane voltage at the onset of a current step, as well as pulse amplitude. For this chopper neuron, spike discharge was evoked from three membrane potentials o f -52, -62, and -74 m V , imposed via constant current injection o f +0.12, 0.00, -0.15 n A , respectively. A. Comparison of voltage traces exhibiting matched sustained voltage levels (star), illustrates the voltage dependence of chopper discharge. Repetitive firing was truncated by hyperpolarization (-74 mV) , while depolarization (-52 mV) resulted in decreased firing rates until onset action potentials were elicited alone. Onset spikes possessed similar timing with changes in membrane potential. B. Changes in chopper discharge may be related to the differential expression of intrinsic membrane properties contributing to firing from these membrane potentials. The increase in T P amplitude upon hyperpolarization o f the holding potential brought the cell to threshold more quickly, resulting in an onset action potential. From membrane potentials below -65 m V , repetitive firing was not observed. T. J. A D A M 194 A T. J. A D A M 195 Figure 5.07 Voltage dependence of chopper neuron intrinsic membrane properties Both subthreshold and suprathreshold chopper membrane behaviour vary with holding potential. For this chopper neuron, holding potentials o f -57, -63, and -70 m V were imposed via +0.09, 0.00, and -0.11 n A respectively. A - C . Voltage responses recorded during 200 ms current pulses demonstrate a greater transient potential following hyperpolarization o f the holding potential. Additionally, the depolarizing voltage sag became less pronounced as the holding potential was increased. D-F. Current-voltage relations for overlying responses illustrate that as the holding potential was shifted negatively, the peak and steady-state relations converged for hyperpolarizations (arrowheads). The sustained trend approached that for peak responses. In depolarizations, the current-voltage relation for peak responses reflected the relatively large T P amplitudes evoked from negative holding potentials (small arrows). T. J. A D A M 196 Current (nA) Current (nA) Current (nA) T. J. A D A M 197 Figure 5.08 Voltage dependence of the depolarizing afterpotential M o s t chopper neurons exhibit a prominent depolarizing afterpotential ( D A P ) upon the offset of a hyperpolarizing stimulus. A . In this exemplar, a mild D A P was exhibited with voltage trajectories returning to levels above -65 mV. The amplitude and onset time course were dependent on the magnitude o f the hyperpolarization. The sustained, depolarizing voltage sag was not prominent in this neuron. B. A D A P was exhibited at the offset of hyperpolarizations regardless o f holding potential in this chopper neuron. From depolarized baselines, the D A P was only visible in voltage traces obtained during large hyperpolarizations. The voltage sag was also pronounced. T. J. A D A M 198 A T. J. A D A M 199 Figure 5.09 Anomalous rectification contributes to the depolarizing afterpotential Blockade of the rapidly and slowly activating anomalous rectifiers have differential effects on chopper neuron behaviour during hyperpolarizations. Voltage responses were evoked from depolarized membrane potentials (-55 raV), above the range o f observed inward rectification. A. Anomalous rectification did not appear to be pronounced in this neuron (arrows). A D A P was also minimal (arrowhead). B. Addit ion of 0.2 m M barium to block rapidly activating anomalous rectifiers produced greater voltage responses to hyperpolarizations, and amplification of the delayed depolarizing voltage sag (arrows). Also , barium increased the amplitude o f the D A P (arrowhead). C. Additional blockade o f slowly activating anomalous rectifiers with 3 m M cesium eliminated the voltage sag during hyperpolarizations. The membrane potential followed a much slower time course to sustained hyperpolarized levels during negative current steps. The D A P was decreased under cesium (arrowhead), and arose later following current offset. D. Peak current-voltage relations derived from voltage traces in A and B demonstrate that barium increased input resistance in the hyperpolarized voltage range. E. Current-voltage relations for the three conditions illustrate the effect o f cesium on sustained responses. The elimination o f the depolarizing voltage sag by cesium application was accompanied by a great increase in input resistance over the same voltage range (< -60 mV). T. J. A D A M 200 T. J. A D A M 201 Figure 5.10 The depolar iz ing afterpotential is mediated i n part by ca lc ium The width and time-course of the depolarizing afterpotential ( D A P ) are affected by reduction in external calcium. The amplitude o f the event is stable. A. Under T T X , a large broad D A P followed the offset of moderate to large negative current pulses. B. Nominal external concentration of calcium (200uM) resulted in a narrower D A P . Offset membrane time constants were not significantly affected by the manipulation. T. J. A D A M 202 B 600 nM TTX 200 uM Calcium 50 ms T. J. A D A M 203 Figure 5.11 Anomalous rectification and a low threshold calcium conductance yield the DAP The depolarizing afterpotential ( D A P ) is eliminated during co-application o f barium, cesium, and nickel. Arrows denote the D A P . A. The D A P was not initially prominent in 600 n M T T X , or in control A C S F (not shown). B. Co-application o f 0.2 m M barium amplified the D A P , but with a short time course. C. Addit ion of 3 m M cesium prolonged the onset time-course, and widened the D A P . D. Co-application of 50 u M nickel almost eliminated the remaining D A P component. T. J. A D A M T. J. A D A M 205 Figure 5.12 A subthreshold sodium conductance contributes to the transient potential The transient potential (TP) has multiple components, one o f which is dependent on sodium flux at membrane potentials subthreshold to action potential generation. A . Control responses evoked from the resting potential (0 m V constant current) showed a minimal T P , and action potential discharge. From hyperpolarized holding potentials, the T P was prominent (E). B . In the presence of 600 n M tetrodotoxin ( T T X ) , spike discharge was eliminated, and depolarizations were reduced. Hyperpolarizations were not affected. C . Current-voltage relations depict reduced depolarization o f the membrane above -45 mV, during sodium channel blockade with T T X . Responses not traversing this voltage level were unaffected by T T X . D . Steady-state I -V relations for this neuron also suggest reduced input resistance in T T X . E. Control voltage responses evoked from negative baselines demonstrated a prominent T P , which reached spike threshold with a minimal subthreshold voltage range. F. Applicat ion of T T X eliminated action potentials and revealed the underlying reduced TP . G. Peak I -V relations show reduced T P amplitude for chopper neurons under T T X . Hyperpolarizations and depolarizations under -45 m V were not affected. H . Sustained I -V relations demonstrate no significant contribution o f T T X sensitive rectification to steady-state responses in this neuron. T. J. A D A M 206 Control B -120-1 i 1 1 r -0.4 -0.2 0.0 0.2 Current (nA) Control r 0.4 1 -0.4 i 1 r -0.2 0.0 0.2 Current (nA) r 0.4 600 nM TTX D -40 n -60-^ -80-^ 9 | -iooH -120-Steady state e - Control • 600 n M T T X — , , 1 1 r --0.4 -0.2 0.0 0.2 0.4 Current (nA) 600 nM TTX H 9 <L> SP > -40--60 H -80 H -100H -120-1 Steady state -0 - Control • TTX T -0.4 -0.2 0.0 0.2 Current (nA) 0.4 T . J .ADAM 207 Figure 5.13 A low threshold calcium conductance contributes to the transient potential The amplitude o f the transient potential (TP) is reduced during nickel blockade of a low threshold calcium dependent conductance. A. The T P was clearly distinguishable in the presence o f T T X at the onset of depolarizations evoked from negative holding potentials. B. Applicat ion of nickel reduced the amplitude o f the T P , but did not affect its time-course. C. Graphically, the effect o f nickel on the amplitude o f the T P is evident in peak relations, while steady-state responses were conserved. T. J. A D A M 208 T. J. ADAM 209 Figure 5.14 The low threshold calcium conductance contributes to the onset spike Blockade of the low threshold calcium conductance with 50 uM nickel further supports a contribution to the transient potential (TP). A. The TP was prominent in voltage traces evoked from holding potentials below -65 mV, and played a role in the determination of onset spike timing. B . With application of nickel, the TP was decreased in amplitude, and spike threshold was raised. Sustained subthreshold responses were not affected. C. A plot of TP amplitude versus current demonstrates nickel's effect on TP amplitude. T. J. A D A M 210 Current (nA) Current (nA) T . J . A D A M 211 Figure 5.15 Anomalous rectification contributes to the nickel sensitive transient potential In addition to a low threshold calcium conductance and a subthreshold sodium conductance, anomalous rectification contributes to the amplitude and timing o f the transient potential (TP). A. The transient potential (TP) was visible in the presence of 600 n M tetrodotoxin ( T T X ) , near the onset o f depolarizations. B. Wi th the addition of 0.2 m M barium, the T P was amplified. However, the onset of the T P was still locked to response onset (arrow). C. Addi t ion of 3 m M cesium resulted in larger TPs, which occurred at a delay dependent on depolarization amplitude (arrows). D. Co-application of 50 u M nickel eliminated the T P . E. Barium application led to T P amplification positive to -75 mV. F. Nickel blocked the T P visible under barium and cesium complication. G. The latency to the peak o f the T P is clearly dispersed over a wider range under cesium application. With increasing current, both latency functions decreased along a single exponential function. T. J. A D A M 212 0.0 0.2 0.4 0.6 0.8 1.0 s 0.0 0.4 0.8 1.2 Current (nA) Current (nA) J 8 0 -1 6 0 i o 40 -cu ™ 20 H i - o -L 0.0 TP latency 0.2 mM Barium 3 mM Cesium 1 1 1 0.4 0.8 Current (nA) "1 1.2 T. J. A D A M 213 Figure 5.16 A-type potassium conductances sculpt the decay of the transient potential Applicat ion of 4 A P in the presence of 600 n M tetrodotoxin ( T T X ) reveals that the T P is partially shaped by a transient A- type potassium conductance, which contributes to the decay o f the T P . A . In T T X , the T P was followed by a transient outward rectification, visible during depolarizations evoked from negative holding potentials (arrow). Application o f 50 u M 4 A P reduced the transient outward sag, and increased the amplitude of the T P . Sustained responses were not significantly affected by 4 A P at this concentration. B. From depolarized baselines, the transient voltage sag was less pronounced; as was the TP . Addi t ion of 4 A P eliminated the sag, and increased the depolarization amplitude for the early period o f the response. Sustained responses were not affected. C . From hyperpolarized holding potentials, current-voltage relations for the transient voltage sag (arrow in A ) reveal its reduction in 50 u M 4 A P . D. Steady state I-V relations were unaffected by 4 A P at this concentration. T. J. A D A M 214 600 n M T T X 50 u M 4 A P + 0.0 0.4 0.8 1.2 0.0 0.4 0.8 1.2 Current (nA) Current (nA) T. J . A D A M 215 Figure 5.17 A-type potassium conductances also contribute to steady-state depolarizations Application of higher concentrations o f 4 A P block not only the transient hyperpolarizing voltage sag observed in Figure 5.16, but also increase sustained depolarizations for 3 2 0 ms stimuli. A . From hyperpolarized membrane potentials, 1 m M 4 A P resulted in greater sustained depolarizations as well as peak responses. B . The effect of 4 A P on steady-state voltage level was also evident in responses evoked from depolarized holding potentials. T. J. A D A M 216 T. J. A D A M 217 Figure 5.18 Other, sustained potassium conductances limit depolarizations Blockade of potassium conductances sensitive to T E A reveal dosage dependent effects o f these conductances on depolarizations. A . In the presence o f 600 n M T T X , action potentials were not expressed in chopper neurons. A small transient potential (TP) was visible at the onset of depolarizing voltage responses for this neuron. B. Co-application of 5 m M T E A increased both transient and sustained depolarizations in chopper neurons, obscuring the TP . C . A t 20 m M , T E A revealed two depolarizing events: a low threshold T P (arrow), and high threshold plateau event (HTS, arrow) that arose from the peak of the T P , at a distinct inflection point (star). D. Current-voltage relations for the peak response support a dosage dependent effect of T E A above 5 m M . Slope resistances were increased under T E A for all depolarizations. E. Sustained current-voltage relations resemble those for transient responses. T. J. A D A M 218 T. J. A D A M 219 Figure 5.19 Chopper neurons exhibit slowly inactivating high threshold calcium spikes The basis for the high threshold spike (HTS) is the high threshold calcium conductance, as revealed by its cadmium sensitivity. A . Under 600 n M T T X , a T P was evident at the onset of voltage responses. B . Applicat ion of 20 m M T E A revealed the high threshold plateau. C. Co-application of cadmium blocked the H T S without ehminating the TP . T. J. ADAM 2 2 0 T. J. A D A M 221 Figure 5.20 A-type potassium conductances truncate chopper responses Chopper spike discharge is sensitive to low levels (50 u M ) o f 4 A P in the external milieu. A . Spike threshold was apparently lowered with 4 A P . Firing was repetitive, and action potentials were relatively wide. In addition, a depolarizing hump followed the onset action potential (only) during voltage responses evoked from hyperpolarized holding potentials. B. From depolarized baselines, the hump was no longer apparent, but other affects on action potentials discharge remained. Spikes were wide, apparently through blockade of an early component of the A H P . Firing was repetitive, and occurred throughout the current pulse. T . J. A D A M 222 HP = -74 mV > S o I CN 0.56 nA 0.48 nA 0.40 nA 0.56 nA 0.48 nA 0.40 nA HP = -52 mV 0.32 nA 0.24 nA 0.16 nA T. J. A D A M 223 Figure 5.21 Sustained potassium conductances contribute to chopper discharge transience Potassium conductances sensitive to T E A also mediate discharge transience, but through slower potassium conductances. A . From hyperpolarized holding potentials, only an onset action potential was evoked during depolarizations. B. A t 5 m M , T E A resulted in repetitive spike discharge during moderate to large current pulses. The second action potential occurred at a pronounced delay from the onset spike. A s the current amplitude was increased, this delay shortened, until obscured by repetitive discharge. During spike trains, action potentials became progressively wider. C. A t 20 m M , the high threshold plateau event (HTS) was revealed. A transient, TEA-insensitive voltage sag followed the sodium dependent onset spike. This was likely the 4 A P sensitive component of the depolarization activated outward rectification of chopper neurons. T. J. A D A M 224 C 20 mM TEA T. J. A D A M 225 Figure 5.22 Intrinsic membrane properties mediate chopper sensitivity to membrane potential The chopper firing pattern is compromised by pre-polarization o f the membrane potential. Current prepulses were delivered immediately prior to a constant suprathreshold test pulse, which evoked typical chopper discharge. A . Voltage responses during prepulses and test pulses, shown overlapped, indicate that the chopper response was disrupted for deviation o f the membrane potential from rest during prepulses. The onset spike remained well-timed, while the timing o f repetitive firing was compromised by the introduction of a delay. This delay was introduced via a transient hyperpolarizing voltage sag that became prominent in test pulses following negative prepulses. A s this event grew, it truncated the first voltage oscillation following the onset spike. From depolarized voltages, prepulses evoked a sustained hyperpolarization-activated voltage sag, which contributed to the T P and subsequent depolarization-activated voltage sag. B. From hyperpolarized baselines, prepulses produced more anomalous rectification, and larger TPs. Onset action potential latency was consequently stable for variant prepulse amplitudes and durations. Repetitive discharge was not observed, regardless of prepulse duration. Following the onset spike, the transient voltage sag was observed, and firing terminated. Sustained voltage level settled to a constant level. C . The onset spike latency decreased slightly with negative prepulses. D. The first interspike interval was lengthened by small pre-polarizations in either the negative or positive direction. It was shortest for chopper responses evoked from rest (0 n A prepulse). E . Spike count decreased with in chopper responses following pre-polarizations from rest. T. J. A D A M 226 Prepulse voltage (mV) Prepulse voltage (mV) Prepulse voltage (mV) T. J. A D A M 227 Figure 5.23 Synaptic depression in inhibitory postsynaptic potentials Inhibitory postsynaptic potentials (IPSPs) exhibit synaptic depression following repetitive electrical stimulation of the primary afferent input fiber tract medial to the L S O . A . Inhibitory synaptic potentials decreased in amplitude after the adapting stimulus train. B. Plotting IPSP amplitude against probe delay from the stimulus train for a typical L S O neuron depicts the time-course of synaptic depression. IPSPs remained depressed in amplitude for approximately 100 ms. T. J. A D A M 228 10 pulses @ 200 per sec. 100 ms Probe delay — 4 16 ms 32 ms 64 ms 128 ms 256 ms 512 ms ~ v — B I P S P Ampl i tude Recovery from Adapta t ion 3 120 100 80 4-6 £ 60 ^ O OH e 20 40 4-10 100 1000 Probe Delay (ms) T. J. A D A M 229 Figure 5.24 Excitatory postsynaptic potentials are stable in LSO neurons Excitatory postsynaptic potentials (EPSPs) express no depression during repetitive electrical stimulation of the primary afferent input fiber tract lateral to the L S O . A. Changes in E P S P amplitude were not evident in chopper neurons tested with short stimulus trains followed by a single probe stimulus. B. Graphical depiction of EPSP amplitude versus probe delay from the stimulus train also demonstrates the stability of EPSPs during repetitive stimulation. The amplitude o f EPSPs remained constant following short stimulus trains. T. J. A D A M 230 10 pulses @ 200 per sec. 1 > g o t 100 ms Probe delay 8 ms 48 ms 88 ms 128 ms 168 ms | 208 ms T3 "a. "e3 £ .S Is 140 120 100 80 60 40 20 0 I EPSP Amplitude Recovery from Synaptic Depression 10 100 1000 Probe Delay T . J . A D A M 231 Figure 6.01 Intrinsic properties of delay neurons in the LSO A second response class ( delay neurons) coexists with chopper neurons in the L S O . Spike discharge and subthreshold responses are qualitatively different from those of chopper neurons. A single representative of the delay neuron class is depicted. A . Delay neurons elicited delayed repetitive firing. The delay varied with the slope o f a depolarizing ramp preceding action potential discharge (arrows). Only large amplitude current pulses evoked an action potential at stimulus onset, superpositioned on a small and brief depolarizing potential. B. Responses evoked by membrane hyperpolarization were passive. N o time- or voltage dependent rectification was apparent in voltage responses. Subthreshold positive current pulses evoked a small and brief depolarizing potential (star), terminating with a transient repolarization, followed by a depolarizing ramp (arrow). Superposition of the reversed hyperpolarizing voltage trace over the depolarizing response of identical magnitude (grey trace, 0.6 nA) illustrates the time course of the transient voltage sag. C . The current-voltage relation is linear for hyperpolarizations, but depicts an early transient outward rectification during depolarizations measured at 10 ms (arrowheads). T. J. A D A M 232 T. J. A D A M 233 Figure 6.02 Temporal pattern of spike discharge for delay neurons Peristimulus time histograms and interspike interval histograms display the temporal discharge pattern o f the delay neuron illustrated in Figure 5.11. Three stimulus amplitudes are depicted. A . The peristimulus time histograms were flat for repetitive firing at all stimulus amplitudes. The early onset spike elicited during large current steps was locked in time to the onset of the stimulus (filled arrow). B. Interspike interval histograms were uni-modal and narrow at low stimulus amplitudes. A t higher current amplitudes, a second wide mode (star) appeared in the histogram, corresponding to the interval between the onset action potential and repetitive firing. Min ima l positive skew indicates little or no spike rate accommodation in delay neurons. C . A dot raster display corresponding to the peristimulus time histogram above illustrates the timing of individual action potentials during repeated trials. The onset spike showed little temporal jitter, while repetitive firing had a variable onset time, relative to the stimulus. T. J. A D A M 234 A. P S T H s 50 40 130 o • | 2 0 10 0 50 40-1 30-o •S 20-10-0 -0.9 nA i i i \ | i i i i r r r r r - p 0 50 1.1 nA B. IS IHs 20 H j 0.9 nA 610-3 100 150 200 Spike latency (ms) u i i • • • 1 • " T ' T ' j 1 1 I r"!" 1 1 1 1 j 50 100 Interval (ms) 20 T 1.1 nA o U 1 0 a S 50 100 150 200 Spike latency (ms) 1 I 1 I I I I I I I I I 1 ! 1 j I I I I j 50 100 Interval (ms) 50: 4 0 : I 2 0 ; 10] 1.3 nA 2 0 11.3 nA 50 ;Too r''^ r ri50 r r'''TT1'"'200 ° Spike latency (ms) . . « .» / ' .•• .'• . > . .•' 4' \'. V; s." s'..' • . > •'• •'• '. ' . •: 1 -A-2 J £ * * 100 Interval (ms) T. J. A D A M 235 Figure 6.03 Spike discharge statistics for delay neurons Delay neuron discharge statistics are distinct from those o f chopper neurons. Data were obtained from one neuron, with the exception of B . A . Intervals following each spike, and variability (error bar = standard deviation) are plotted against latency. Stability of firing rate (interval) and timing precision (error bars) were apparent at all stimulus amplitudes. In addition, an additional long and variable first interval was produced by the evokation of an onset spike. B. Onset spike latency and variability decreased with current level (different neuron). C . The coefficient of variation was above 0.2 for the first interval, but is stable and minimal (0.1) during repetitive firing. D. Rate accommodation was minimal; spike rate decreasing to no less than 80 % o f the first interval in repetitive firing over the 200 ms stimulus period. E . Instantaneous spike rate increased approximately linearly with current. T. J. A D A M 236 Interval-Latency Relation 0.8 nA 0.9 nA 1.0 nA 1.1 nA 1.2 nA 1.3 nA Mean latency (ms) 150 200 Discharge Regularity 0.8 nA 0.9 nA 1.0 nA 1.1 nA 1.2 nA 1.3 nA T 0.8 1.0 1.2 Current (nA) Rate Accomodation E "H 100-o o u 80-60-0.8 nA 0.9 nA 1.0 nA 1.1 nA 1.2 nA 1.3 nA g40-a 3 20-T—[— i— i—p—i—j—i—i—i— i—|—r 50 10 150 Latency (ms) 200 4) •oo t I OH -1—I—1—I—I—1—I—I—I—I—I—I—I—I—I—I— 50 100 150 M e a n latency (ms) Input-Output Function Rate 200 r 10 8-g CD 4 ^ o o 2 a - i — i — i — i — i — r 0.6 0.8 1.0 1.2 1.4 1.6 Current (nA) 1.8 2.0 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0089656/manifest

Comment

Related Items