| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Neurobiology, Northeastern Ohio Universities College of Medicine, Rootstown, Ohio
Submitted 12 September 2006; accepted in final form 22 November 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). Therefore it is important to understand how the auditory system extracts temporal information from complex sounds.
Numerous studies using FM sounds have shown that the temporal structure of FM is reflected by neuronal responses in the auditory nerve and in lower auditory brain stem nuclei. At progressively higher levels of the auditory system, neurons become more selective for FM (review: Langner 1992). The inferior colliculus (IC), a large midbrain auditory processing center in vertebrates, contains neurons that are tuned to narrow ranges of modulation frequencies (Casseday et al. 1997; Langner and Schreiner 1988; Rees and Møller 1983; Schuller 1979).
The first systematic study of FM response selectivity in IC neurons was performed in Myotis lucifugus (Suga 1968). Suga found a number of neurons that responded only to FM sweeps but not to pure tones. Similar results were reported later in the IC of other bats (Casseday and Covey 1992; Fuzessery 1994) and rats (Poon et al. 1992).
A few possible neural mechanisms that can underlie FM rate and direction tuning in IC neurons of mustached and pallid bats were previously proposed (Brimijoin and O'Neill 2005; Fuzessery et al. 2006; Heil et al. 1992; Rall 1964; Suga 1965; Suga and Schlegel 1973). The sideband inhibition model proposes that a directionally selective neuron receives two inputs, one excitatory and the other inhibitory. Typically the best frequencies of these opposing inputs are different. Sweeps in the preferred direction will elicit stronger responses because they stimulate the excitatory area before the inhibitory (Brimijoin and O'Neill 2005; Fuzessery et al. 2006; Heil et al. 1992; Rall 1964; Suga and Schlegel 1973). By contrast, sweeps in the nonpreferred direction elicit weaker responses because they activate the inhibitory area before the excitatory. Another model (serial excitation model) is based on temporally offset, subthreshold excitatory inputs that are anatomically ordered by best frequency along a dendrite (Rall 1964). An FM sweep that stimulates the inputs in a sequence toward the soma results in summation, whereas stimulation in a sequence away from the soma does not. Thus FM selectivity is likely to be a result of the spectrotemporal interaction of excitatory and inhibitory inputs on IC neurons.
IC neurons also showed response selectivity for a number of other stimulus parameters such as interaural intensity difference (Klug et al. 1995; Park and Pollak 1993, 1994), sound duration (Casseday et al. 1994; Covey et al. 1996), sound frequency, and sound level (Faingold et al. 1989; Le Beau et al. 1996; Pollak and Park 1993; Vater et al. 1992). These types of response selectivity further suggest that spectrotemporal integrations at the level of the auditory midbrain might (at least partially) be responsible for the response selectivity observed in IC neurons. At present little is known about postsynaptic mechanisms underlying this integration.
Our goal was to study postsynaptic mechanisms underlying spectrotemporal integration in the IC. Intracellular responses were recorded in IC of the little brown bat in response to different FM sweeps. Bats have proved to be excellent models for studying the processing of FM sounds because these sounds are the most common signals that bats use for echolocation and communication (Simmons and Stein 1980; Simmons et al. 1979).
We found that a majority of IC neurons show postsynaptic responses to a wide range of sound frequencies. The most common response pattern to FM sweep is inhibitory postsynaptic potential (IPSP)–excitatory (E)PSP(spike)–IPSP. About one third of IC neurons perform nearly linear temporal summation across a wide range of sound frequencies. This study shows that mechanisms of spectral integration are important for complex sound processing in the IC.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Experimental subjects comprised 22 little brown bats, Myotis lucifugus. For surgery, the animal was anesthetized by isoflurane inhalation (1.5–2.0%, isoflurane administered by a precision vaporizer). After incision of the skin and clearing of the tissues above the skull, a small metal rod was glued to the skull using glass ionomer cement. After the surgery, animals were allowed to recover for 2–3 days in individual holding cages.
Recordings were made from awake bats. During the recording session the animal was placed inside a single-walled sound-attenuating chamber (Industrial Acoustics). The metal rod on the bat's head was secured to a small holder for restraining the animal's head atraumatically, leaving the ears unobstructed for free-field acoustic stimulation. A small hole (about 50 µm) was then made in the skull overlying the IC through which a recording electrode was inserted to reach the IC. Throughout the recording session, the animal was offered drinking water periodically and monitored for signs of discomfort. After a recording session of 6–8 h, the exposed skull was covered with sterile bone wax and the animal was returned to its holding cage. Such experiments proceeded every 2–3 days for a maximum of 2 wk.
Procedures used in this study were approved by the Institutional Animal Care and Use Committees at the Northeastern Ohio Universities College of Medicine.
Acoustic stimulation
Acoustic stimuli, consisting of linear downward FM sweeps, were delivered to the bat by a free-field ultrasonic loudspeaker (Ultra Sound Advice US-LS) located 30 cm in front of the bat. The outputs of the loudspeaker were measured with a 1/4-in. microphone (Brüel & Kjær 4135) and found to be flat ±6 dB between 20 and 80 kHz, the frequency range used in the experiments. The parameters of the acoustic stimuli were controlled by D/A hardware and software from Tucker-Davis Technologies (TDT System III) with sampling rate 200 kHz. FM stimuli that mimicked the bats' natural FM signal were swept from 80 to 20 kHz in 4 ms (including 0.25-ms rise/fall time) and were termed the entire FM sweep, which was also arbitrarily divided into three equal 20-kHz subcomponents (80–60 kHz, 60–40 kHz, and 40–20 kHz). Each subcomponent had 1.33-ms duration including 0.25-ms rise/fall time. All FM sweeps were presented at a wide range of sound levels, from 0 to 80 dB SPL. Stimuli were delivered at a rate of four pulses per second. Unless otherwise specified, for each recorded neuron we presented four different types of FM sweeps (entire FM sweep and three FM subcomponents) at 0, 20, 40, 60, and 80 dB SPL. The entire protocol was repeated as many times as possible while the neuron showed a stable membrane resting potential (fluctuations of the baseline did not exceed the range of a few millivolts). Only neurons that were tested at least three times with the entire protocol were included in our data analysis.
Recording procedure
Microelectrodes were made from 1.2-mm-diameter quartz glass (Sutter Instruments, Novato, CA) filled with 3 M potassium acetate. Micropipettes, with impedance between 50 and 120 M
were pulled on a Flaming-Brown micropipette puller (Sutter P2000). The electrode was positioned above the IC by means of a precision digital micromanipulator and lowered to the dorsal brain surface. The relative position of each electrode was monitored from the readouts of digital micrometers using a common reference on the skull. Vertical advancement of the electrode was made by a precision microdrive in 2- to 3-µm steps (KOPF Model 660) from outside the sound-attenuating chamber. After placement of the electrode on the surface of the IC using a surgical microscope (Leica MZ9.5), the exposure was filled with 4% agar.
Intracellular responses of IC neurons were amplified through a single-channel amplifier (model IR183A, Cygnus Technology) and monitored on a digital oscilloscope (Yokogawa DL1640). Intracellular waveforms from IR183A and sound stimuli from the TDT system were digitized and then stored on a computer hard drive using EPC-10 digital interface and PULSE software from HEKA at a bandwidth of 100 kHz. To detect cell impalement, small (5- to 100-nA) current pulses of 100-ms duration were delivered through the microelectrode. A sudden negative DC shift and the presence of synaptic potentials indicated an intracellular impalement, which was often verified by passing positive current to evoke action potentials. Stable intracellular impalements were signaled by a prolonged (>3 min), stable drop (>40 mV) in the DC level. Intracellular recordings typically lasted 3–5 min (maximum 40 min). During intracellular recording the cell membrane resting potential usually fluctuated not more than a few millivolts. Successful intracellular recording was always accompanied by the presence of postsynaptic potentials (PSPs) and in almost all our recordings by large action potentials (>30 mV). Each of our recorded cells satisfied all of the above criteria for being studied with our experimental protocol. A sudden decrease in membrane resting potential was typical just before a recorded neuron was lost. Intracellular recordings were performed for IC cells located at depths of 200 to 1,200 µm primarily to collect responses from cells in the central nucleus of the IC. Different angles of electrode penetrations were used in different experiments for the same animal to avoid local damage of the IC. Therefore recording depths indicated for neurons presented in this study do not give an exact location of a recorded neuron in the IC along the dorsoventral axis.
Data analysis
During intracellular recording we often observed response variability from stimulus presentation to another. Therefore an algebraic average for all traces recorded at each stimulus parameter was calculated (Fig. 1). A baseline value was first calculated by determining the mean value for all samples of each waveform for a time window 
200 ms (sampling rate = 10 µs). Because spikes and/or large PSPs could alter the mean value we eliminated all values that exceeded 1 SD from that mean. Then a new baseline mean value was calculated. PSPs were defined as depolarizing or hyperpolarizing fluctuations on the averaged traces that exceeded 2 SDs (95% confidence limits) from the baseline and occurring within a 50-ms time window after stimulus onset. PSPs latencies were calculated to crossing of the 2 SDs threshold.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Intracellular responses of 117 IC neurons to FM stimuli (80–20 kHz, 4-ms duration; 0.25-ms rise/fall time) were recorded in 22 awake little brown bats.
Neurons in the IC displayed a variety of synaptic responses to acoustic stimulation that reflected both excitatory and inhibitory inputs. The most common response pattern (73/117 units or 62%) was hyperpolarization followed by depolarization with or without spike followed by hyperpolarization [IPSP–EPSP(spike)–IPSP]. Only two neurons from this population did not exhibit any spikes in their responses. Typical responses contained one or two spikes (67/73), but much less often three or four action potentials (4/73) were produced. The sound level did not change the general response pattern of these neurons; depolarization was preceded and followed by hyperpolarization with or without spikes.
Intracellular responses of a representative neuron from this population are shown in Fig. 2, A–D. In response to the 80- to 20-kHz FM sweep, this unit exhibited a relatively stable response pattern [IPSP–EPSP(spike)–IPSP], which was independent of sound level. In response to the FM sweep presented at a subthreshold level for spikes (10 dB SPL) a small initial IPSP occurred followed by an EPSP (Fig. 2A), whereas at suprathreshold sound levels (40 and 60 dB SPL) this unit exhibited the IPSP–EPSP–spike–IPSP response pattern (Fig. 2, B and C). In the waveform shown in Fig. 2C an action potential was generated at a level below the resting membrane potential. In spite of this, we believe that this spike was generated by an EPSP that did not reach the resting membrane voltage because it occurred during an IPSP. Responses of this neuron to lower and higher sound levels showed clear depolarization (Fig. 2, B and D). At the sound level of 80 dB SPL this neuron did not discharge but it retained the IPSP–EPSP–IPSP response pattern (Fig. 2D). Thus the vast majority of IC neurons in response to FM sweeps showed IPSP–EPSP–IPSP response pattern with one or two spikes.
|
The response pattern of the remaining 19/117 neurons (17%) could not be classified according to any of the two categories of IC neurons described above. Nine of them responded with IPSPs only (see Fig. 5 for reference). The remaining 10 neurons showed either an IPSP–EPSP–spike (three units) or EPSP–spike–IPSP response pattern (four units) or their response pattern was unstable and varied from one presentation to another (three units).
|
As described above, 62% of IC neurons responded to the 80- to 20-kHz FM sweep with complex multicomponent postsynaptic responses: IPSP–EPSP(spike)–IPSP. This response pattern may be evoked by convergence of different frequency-dependent inputs to each IC neuron. Furthermore, these inputs should arrive within a limited time window to be integrated. We hypothesized that complex multicomponent postsynaptic responses in the IC are the result of spectrotemporal integration. To test this hypothesis we compared postsynaptic responses of IC neurons to the entire FM sweep and to each of three arbitrarily but equally divided FM subcomponents: 80–60, 60–40, and 40–20 kHz (1.33-ms duration; 0.25-ms rise/fall time). Unless otherwise specified, all FM sweeps were presented at the same sound level, usually at 10 or 20 dB above a unit's threshold for spikes.
Spectral integration
Intracellular responses of 41 of 73 IC neurons exhibiting the IPSP–EPSP(spike)–IPSP were studied in response to the entire FM sweep and also to three FM subcomponents. Postsynaptic responses (PSPs with or without spikes) in response to the first (80- to 60-kHz), second (60- to 40-kHz), and third (40- to 20-kHz) FM subcomponents were evident in 13 of 41 neurons. Postsynaptic responses to the first (80- to 60-kHz) and the second (60- to 40-kHz) FM subcomponents were displayed by 24 of 41 neurons. Two of 41 neurons showed postsynaptic responses to the second (60- to 40-kHz) and third (40- to 20-kHz) FM subcomponents. Finally, two of 41 neurons showed postsynaptic responses to the first (80- to 60-kHz) FM subcomponent only. Thus a vast majority of IC neurons (39/41) responded to at least two FM subcomponents, corresponding to the frequency range of 40 kHz.
Action potential response thresholds to the entire FM sweep were often different from the thresholds to individual FM subcomponents. Twenty-six percent (11 of 41) of units fired spikes in response to the entire FM sweep, whereas they showed only PSPs (excitatory and/or inhibitory) without spikes in response to FM subcomponents presented at the same sound level (Fig. 3). Data from the unit shown in Fig. 3 demonstrate a complex response pattern (IPSP–EPSP–spike–IPSP) to stimulation with the entire FM sweep (Fig. 3A). Truncating the signal to the initial 80- to 60-kHz FM and the second 60- to 40-kHz FM subcomponents resulted in a postsynaptic response pattern similar to that received for the entire FM sweep, but below the unit's spike threshold (Fig. 3, B and C). In response to the 40- to 20-kHz FM it showed a small EPSP followed by a small IPSP (Fig. 3D).
|
Temporal integration
The hypothesis that temporal summation across different frequency bands can produce the response to the entire FM sweep was tested in 16 IC neurons. For each neuron, the algebraic sum of all postsynaptic responses to the three FM subcomponents was compared with its response to the entire FM sweep. To sum the individual PSPs, responses to the FM subcomponents were realigned to account for their position within the entire FM sweep. In other words, before summation the PSP responses to the second and third FM subcomponents were shifted +1.33 and +2.66 ms, respectively. The summated potential then was normalized to the unit's resting membrane potential. A correlation coefficient was used to determine how well the summated potentials matched unit responses to the entire FM sweep. Because all neural responses were collected with the same sampling rate (100 kHz), we were able to calculate correlation coefficients between the event amplitudes using a high-resolution data set (within a 50-ms time window beginning from stimulus onset 5,000 points were compared between responses). Action potentials could potentially contaminate the correlation coefficient values. Therefore data points during action potentials (1.5 ms) and correspondent points on the other trace were excluded from analysis.
For all 16 IC neurons studied during these experiments, the correlation values were calculated for responses measured at sound levels 10–20 dB above each unit's threshold for action potentials. Five of 16 neurons had correlation coefficient values between 0.6 and 0.85. A representative neuron from this population is shown in Fig. 4. All stimuli for this neuron were presented at a sound level 20 dB above action potential threshold. The response to the first FM subcomponent was similar to that elicited by the entire FM sweep (compare Fig. 4, A and B). In response to the second FM subcomponent (60–40 kHz) this neuron showed a bimodal IPSP containing a maximum point of depolarization in the middle (Fig. 4C). Finally, in response to the 40- to 20-kHz FM subcomponent this neuron showed an EPSP (Fig. 4D). The algebraic sum of postsynaptic responses to the three FM subcomponents correlated well (r = 0.69, P < 0.001) with the unit's response to the entire FM sweep (Fig. 4, bottom).
|
For the remaining 11/16 units the correlation values were either low (5/16 units; from 0.3 to 0.6) or close to zero (6/16 units; from 0 to <0.3).
Thus one third of the IC neurons performed temporal summation across a wide range of sound frequencies in a nearly linear manner. The remaining two thirds of neurons performed nonlinear temporal summation with different degrees of nonlinearity.
Sound level and temporal summation
The influence of sound level on postsynaptic responses was studied in 14 IC neurons. For about one third of these neurons (5/14) summated postsynaptic potentials in response to three FM subcomponents were highly correlated with responses to the entire FM sweep at sound levels from 10 to 20 dB above the unit's threshold for spikes. At high sound levels (60, 80, or 90 dB SPL), however, this correlation was substantially reduced.
Postsynaptic responses of a representative neuron from this population are shown in Fig. 6. This figure demonstrates the effect of sound level on the relationship between postsynaptic potentials in response to the entire FM sweep and the summated response to the FM components. The correlation coefficient values are shown for five different sound levels (0, 20, 40, 60, and 80 dB SPL). Analysis of the data reveals a low correlation between the summated PSPs and the response to the entire FM sweep at 0 dB SPL (see Fig. 6B for the reference). Poor correlations at 0 dB SPL were expected because this neuron showed very little if any response at this sound level. At 20 dB SPL this neuron showed a subthreshold postsynaptic response to the entire FM sweep. At this sound level the correlation coefficient was maximal (see Fig. 6C for the reference). When the sound level was further increased to 40, 60, and 80 dB SPL correlation coefficients declined to 0.5, 0.3, and 0.07, respectively (Fig. 6A; see Fig. 6, D, E, and F for reference).
|
|
Our data demonstrate that the vast majority of IC neurons (39 of 41) responded to more than one FM subcomponent, suggesting that these neurons integrate information across a wide range of sound frequencies. If convergence occurs at the IC, the timing of these inputs to IC neurons would be important for defining the response to FM stimuli. To address this question we studied intracellular responses of eight IC neurons when the three FM subcomponents were presented at different intercomponent time delays (0, 1, 2, 4, 8, 16, and 32 ms). All FM subcomponents were always presented in the same sequence in which they appear within the entire FM sweep. One limitation to this approach is that each of the three FM subcomponents has its own rise/fall time of 0.25 ms. In other words, when all three FM subcomponents are combined in a train with 0-ms intercomponent time intervals the result is a 4-ms FM sweep (the entire FM sweep) with two embedded notches. Such a stimulus has less energy than that of the entire FM sweep and a different spectrum. Therefore to determine how these differences might modify neurons' response patterns, responses from all eight IC neurons to the entire FM sweep were compared with responses to a train of three FM subcomponents with 0-ms intersubcomponent time interval. For all of these neurons we found little differences between responses to these two types of FM stimuli (Fig. 8).
|
|
The remaining five of eight neurons were affected differently as the delay between FM subcomponents was increased. For three, the response to the entire FM sweep was mainly determined by the response to the first FM subcomponent. In other words, they showed nearly identical responses to the entire FM sweep and to the first FM subcomponent. Therefore increasing the delay between FM subcomponents did not alter their response patterns.
The other two of five neurons fired spikes to the entire FM sweep, but displayed subthreshold PSPs in response to individual FM subcomponents presented alone. Increasing the intersubcomponent delay increased the interval between these PSPs without their modification, suggesting that firing of these neurons was a result of temporal summation across a wide range of sound frequencies.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Methodological issues
Intracellular recording from awake animals is a technically challenging procedure that results in very short recording time and therefore small quantities of data for individual neurons. During our experiments we usually could not hold IC neurons longer than 2–3 min. However, in spite of short recording time, we do believe that the quality of our intracellular recordings is comparable with that of previous intracellular studies. The range of resting membrane potentials and spike amplitudes presented in this study are similar to those reported by others (Kuwada et al. 1997; Nayagam et al. 2005, 2006; Nelson and Erulkar 1963; Pedemonte et al. 1997; Volkov and Galazyuk 1991, 1992). All but two neurons that we report here displayed both synaptic and action potentials. Therefore we feel confident that our intracellular data are valid to determine postsynaptic mechanisms underlying FM sound processing in the inferior colliculus of awake bats.
Temporal summation of postsynaptic events
Previous studies suggest that integration of information across multiple frequency bands plays a crucial role in processing of complex sounds both in echolocating bats (Fuzessery and Hall 1996; Fuzessery et al. 2006; O'Neill and Brimijoin 2002; Portfors and Wenstrup 2001; Wenstrup and Leroy 2001) and in nonecholocating animals (Linden et al. 2003; Phillips and Irvine 1981; Portfors and Felix 2005; Sutter and Schreiner 1991). Linear summation of inhibitory and excitatory events was previously shown to play a major role in cellular mechanisms responsible for response selectivity to the direction of visual motion in neurons of the cat primary visual cortex (Jagadeesh et al. 1993; Priebe and Ferster 2005). Our study supports these conclusions. We found that one third of IC neurons can integrate information using nearly linear summation.
In this study we demonstrated that the majority of IC neurons (39 of 41) receive inputs over a wide range of sound frequencies at sound levels close to a unit's threshold for spikes because they show postsynaptic responses to two or three FM subcomponents. Summation of these inputs determines neural responses to the entire FM sweep. IC neurons performed either near linear (30% of units) or nonlinear summation (70% of units) of these inputs over time. The result of this summation facilitated or suppressed responses of IC neurons.
It is of interest that the majority of IC neurons (roughly 70%) show nonlinear integration. There are a few potential intrinsic cellular mechanisms that could be involved in such integration. For instance, the hyperpolarization-activated current such as Ih current in auditory brain stem neurons can contribute to the precise analysis of temporal information (Koch and Grothe 2003). This current can either reduce or improve temporal summation of excitatory and inhibitory potentials in auditory neurons (Bal and Oertel 2000; Koch and Grothe 2003; Oertel 1999). Additional evidence of hyperpolarization-activated intrinsic mechanisms comes from both in vitro study of rat brain slices and in vivo recording in the inferior colliculus of unanesthetized rabbits (Sivaramakrishnan et al. 2004). Because the majority of IC neurons in our study showed an initial hyperpolarization in their responses these hyperpolarization-activated intrinsic mechanisms might play a role in creation of complex responses to FM sweeps. Depolarization-activated low-threshold potassium current was recorded in the auditory brain stem neurons (Adam et al. 2001). Low- and high-threshold Ca2+ channels, which can be activated by depolarization, were found in the inferior colliculus neurons of adult rats (N'Gouemo and Morad 2003). Thus in addition to patterns of synaptic input, response properties of the IC neurons can be shaped by the intrinsic properties of the postsynaptic neuron. Future study of the intrinsic properties of IC neurons may clarify the role of these mechanisms in temporal integration.
Excitation surrounded by inhibition
Our data suggest that most IC neurons respond to the entire FM sweep with an IPSP–EPSP(spike)–IPSP response function. Independent of sound level these neurons usually show one or sometimes two action potentials in their responses. Early inhibition in auditory neuron responses to sounds was observed by others (Carney and Yin 1989; Faure et al. 2003; Nelson and Erulkar 1963; Park and Pollak 1993). Both early and late inhibitory components were observed during in vivo patch-clamp recording from the inferior colliculus neurons in big brown bats (Covey et al. 1996). It is unclear whether the early and late inhibitions are two parts of a long-lasting inhibition triggered by FM stimulus onset and interrupted by a strong excitation arriving a few milliseconds later than inhibition. Our data cannot provide a definitive answer to this question. In response to individual FM subcomponents presented alone some neurons showed an IPSP–EPSP(spike)–IPSP response pattern that was similar to that of the entire FM sweep (Figs. 3 and 4). Therefore for these neurons it is difficult to speculate about the origin of the early and late inhibitions for their responses to FM sweeps. Other neurons showed short-latency IPSP in response to one FM subcomponent while exhibiting longer-latency EPSP(spike) or EPSP(spike)–IPSP responses to another (Fig. 9), suggesting that these neurons receive early and late inhibitory inputs from different frequency ranges. Future studies that use intracellular recordings from individual IC neurons in response to both FM sweeps and pure tones presented at different frequencies might clarify this point.
Whether the late IPSP is caused by activation of intrinsic conductances after action potential(s) or after EPSP is unclear. Our data suggest that an intrinsic mechanism is unlikely to be responsible for the late IPSPs. First, in many of our neurons we did not observe spikes while they showed clear IPSP–EPSP–IPSP response patterns for some stimulus conditions. Second, we also often observed spontaneous spikes that were not followed by an IPSP. Nevertheless, we cannot rule out the possibility for an intrinsic source of the late inhibitory response component. Future experiments that use blocking of different intrinsic conductances may elucidate this point.
The fact that inhibition flanks excitation in many IC neurons, thus restricting the time available to response, might have an important behavioral relevance for bats. When a bat pursues a flying insect, its sonar emission rate is increased from five to 20 pulses per second during the search phase and to 100–200 pulses per second just before prey capture (Griffin 1986; Kalko 1995; Schnitzler et al. 1987). It is important for bats to analyze all returning echoes separately from each other when the time interval between them can be as small as a few milliseconds. Therefore it would be very important for these animals to restrict the time of neural response to avoid response overlap. At the same time this mechanism should be level independent because of the dramatic difference between sound level of emitted pulses and returning echoes. Thus a narrow response window for IC neurons should be critical for successful behavioral performance of bats.
Inhibition preceding excitation, also previously reported for nonecholocating animals (Nelson and Erulkar 1963), may play an important role for central auditory processing to control a balance between excitation and inhibition. This balance can vary systematically as a function of stimulus parameters. For example, balance between inhibition and excitation can be altered as a function of sound level, causing response latency to increase. Neurons exhibiting so-called paradoxical latency shift were previously reported both for echolocating bats (Galazyuk and Feng 2001; Galazyuk et al. 2005; Sullivan 1982) and for nonecholocating animals including insects (cats: Rose et al. 1963; gerbils: Klug et al. 2000; insects: Krahe et al. 2002; frogs: Galazyuk et al. 2005).
Response latency and FM response selectivity
Our data demonstrate that some IC neurons showed different PSP latencies in response to different FM subcomponents. Furthermore, the latencies to different FM subcomponents often did not correlate with the sequential order of these subcomponents within the entire FM sweep. Response latencies to the second FM (60–40 kHz) subcomponent (inhibitory input) were a few milliseconds shorter than response latencies to the first subcomponent (80–60 kHz) (excitatory input). Suppression of spikes when excitation to the initial part of the FM sweep was preceded by the inhibition to the later part of the FM sweep is equivalent to backward masking (Zwicker and Fastl 1990). It is unclear what mechanism is responsible for the long-latency excitation or short-latency inhibition. One plausible scenario is that a sustained excitation has the same latency as that of the initial inhibition but the strength of inhibition is greater. Therefore excitation is delayed by this stronger inhibition. This explanation is consistent with extracellular recording studies that show that some IC neuron discharge latency can be shortened by 
-aminobutyric acid (GABA) or glycine receptor antagonist application (Casseday and Covey 1996; Johnson 1993; Park and Pollak 1993). In the IC, inhibition may extend the range of latencies that are larger than would be expected from synaptic delays and axon length (Casseday and Covey 1996; Haplea et al. 1994). Another piece of the evidence comes from study of auditory neurons exhibiting a "paradoxical latency shift" (Galazyuk and Feng 2001; Galazyuk et al. 2005; Sullivan 1982). These neurons show longer response latencies at high sound levels. Application of GABAA blocker to these IC neurons revealed shorter-latency firing in their discharge pattern. In contrast to these findings several studies demonstrated that in the inferior colliculus inhibition has little if any contribution to the lengthening of response latencies (Fuzessery et al. 2003; Le Beau et al. 1996). Further studies are necessary to clarify this point.
Fast inhibitory inputs described in our study might be responsible for the FM sweep rate selectivity. Such neurons would not respond to FM sweeps with long durations (low sweep rate) because a fast inhibition in response to the middle part of the entire FM sweep might then coincide with a late excitation in response to the beginning of the entire FM sweep. This is exactly what happened when we increased intersubcomponent delay in a train of FM subcomponents from 2 to 4 ms. Thus for such neurons fast inhibitory input plays a crucial role in creation of FM sweep response selectivity. A similar mechanism was reported for duration-selective neurons in the inferior colliculus of big brown bats (Faure et al. 2003). The authors presented evidence that inhibition can play a significant role in temporal masking phenomena, including forward, simultaneous, and backward masking.
The importance of GABAergic inhibition in FM sweep direction and rate response selectivities was previously shown for IC neurons (Fuzessery and Hall 1996; Fuzessery et al. 2006). Using a two-tone inhibition paradigm these authors described a few inhibitory mechanisms that can shape selectivity for FM sweep direction and rate. It turns out that two different inhibitory mechanisms can create similar tuning to the rate of FM sweeps. The first mechanism suggests that duration tuning determines rate tuning. In the IC neurons that are not duration tuned another mechanism plays a role for the rate tuning. Delayed high-frequency inhibition at slow FM rates overlaps excitation. The authors suggested that it is possible for both mechanisms to work in concert. The cellular mechanism underlying FM rate selectivity discovered in our study is different from the two mechanisms described above. Our data suggest that delayed high-frequency excitation and short-latency low-frequency inhibition might be responsible for both FM sweep rate and direction selectivity. Thus it further supports a hypothesis that more than a single mechanism can be used by the auditory system to achieve the same endpoint (Fuzessery et al. 2006). A similar conclusion was made after multiple attempts to find a single cellular mechanism that might be responsible for sound duration tuning in the auditory system of different animals (Casseday et al. 1994; Fuzessery and Hall 1999).
In conclusion, the main point we seek to emphasize is that the majority of IC neurons integrate information across a wide range of sound frequencies. Some of these neurons perform nearly linear summation, whereas others exhibit nonlinear summation with different degrees of nonlinearity. This is an important finding because it may shed light on mechanisms by which individual auditory neurons can analyze complex sounds.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: A. V. Galazyuk, Northeastern Ohio Universities College of Medicine, 4209 State Route 44, Rootstown, OH 44272 (E-mail: agalaz{at}neoucom.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Bal R, Oertel D. Hyperpolarization-activated, mixed-cation current (I(h)) in octopus cells of the mammalian cochlear nucleus. J Neurophysiol 84: 806–817, 2000.
Carney LH, Yin TC. Responses of low-frequency cells in the inferior colliculus to interaural time differences of clicks: excitatory and inhibitory components. J Neurophysiol 62: 144–161, 1989.
Casseday JH, Covey E. Frequency tuning properties of neurons in the inferior colliculus of an FM bat. J Comp Neurol 319: 34–50, 1992.[CrossRef][ISI][Medline]
Casseday JH, Covey E. A neuroethological theory of the operation of the inferior colliculus. Brain Behav Evol 47: 311–336, 1996.[ISI][Medline]
Casseday JH, Covey E, Grothe B. Neural selectivity and tuning for sinusoidal frequency modulations in the inferior colliculus of the big brown bat Eptesicus fuscus. J Neurophysiol 77: 1595–1605, 1997.
Casseday JH, Ehrlich D, Covey E. Neural tuning for sound duration: role of inhibitory mechanisms in the inferior colliculus. Science 264: 847–850, 1994.
Covey E, Kauer JA, Casseday JH. Whole-cell patch-clamp recording reveals subthreshold sound-evoked postsynaptic currents in the inferior colliculus of awake bats. J Neurosci 16: 3009–3018, 1996.
Erulkar LD, Butler RA, Gerstein GL. Excitation and inhibition in cochler nucleus. II. Frequency-modulated tones. J Neurophysiol 31: 537–548, 1968.
Faingold CL, Gehlbach G, Caspary DM. On the role of GABA as an inhibitory neurotransmitter in inferior colliculus neurons: iontophoretic studies. Brain Res 500: 302–312, 1989.[CrossRef][ISI][Medline]
Faure PA, Fremouw T, Casseday JH, Covey E. Temporal masking reveals properties of sound-evoked inhibition in duration-tuned neurons of the inferior colliculus. J Neurosci 23: 3052–3065, 2003.
Fuzessery ZM. Response selectivity for multiple dimensions of frequency sweeps in the pallid bat inferior colliculus. J Neurophysiol 72: 1061–1079, 1994.
Fuzessery ZM, Hall JC. Role of GABA in shaping frequency tuning and creating FM sweep selectivity in the inferior colliculus. J Neurophysiol 76: 1059–1073, 1996.
Fuzessery ZM, Richardson MD, Coburn MS. Neural mechanisms underlying selectivity for the velocity and direction of FM sweeps in the inferior colliculus of the pallid bat. J Neurophysiol 96: 1320–1336, 2006.
Fuzessery ZM, Wenstrup JJ, Hall JC, Leroy S. Inhibition has little effect on response latencies in the inferior colliculus. J Assoc Res Otolaryngol 4: 60–73, 2003.[CrossRef][Medline]
Galazyuk AV, Feng AS. Oscillation may play a role in time domain central auditory processing. J Neurosci 21: RC147, 2001.
Galazyuk AV, Lin W, Llano D, Feng AS. Leading inhibition to neural oscillation is important for time-domain processing in the auditory midbrain. J Neurophysiol 94: 314–326, 2005.
Griffin DR. Listening in the Dark. Ithaca, NY: Cornell Univ. Press, 1986.
Haplea S, Covey E, Casseday JH. Frequency tuning and response latencies at three levels in the brainstem of the echolocating bat, Eptesicus fuscus. J Comp Physiol A Sens Neural Behav Physiol 174: 671–683, 1994.
Heil P, Langner G, Scheich H. Processing of frequency-modulated stimuli in the chick auditory cortex analogue: evidence for topographic representations and possible mechanisms of rate and directional sensitivity. J Comp Physiol A Sens Neural Behav Physiol 171: 583–600, 1992.[Medline]
Jagadeesh B, Wheat HS, Ferster D. Linearity of summation of synaptic potentials underlying direction selectivity in simple cells of the cat visual cortex. Science 262: 1901–1904, 1993.
Kalko E. Insect pursuit, prey capture, and echolocation in pipistrelle bats. Anim Behav 50: 861–880, 1995.[CrossRef]
Klug A, Park TJ, Pollak GD. Glycine and GABA influence binaural processing in the inferior colliculus of the mustached bat. J Neurophysiol 74: 1701–1713, 1995.
Krahe R, Budinger E, Ronacher B. Coding of a sexually dimorphic song feature by auditory interneurons of grasshoppers: the role of leading inhibition. J Comp Physiol A Sens Neural Behav Physiol 187: 977–985, 2002.[CrossRef][ISI][Medline]
Langner G. Periodicity coding in the auditory system. Hear Res 60: 115–142, 1992.[CrossRef][ISI][Medline]
Langner G, Schreiner CE. Periodicity coding in the inferior colliculus of the cat. J Neurophysiol 60: 1799–1822, 1988.
Le Beau FE, Rees A, Malmierca MS. Contribution of GABA- and glycine-mediated inhibition to the monaural temporal response properties of neurons in the inferior colliculus. J Neurophysiol 75: 902–919, 1996.
Linden JF, Liu RC, Sahani M, Schreiner CE, Merzenich MM. Spectrotemporal structure of receptive fields in areas AI and AAF of mouse auditory cortex. J Neurophysiol 90: 2660–2675, 2003.
Nayagam DA, Clarey JC, Paolini AG. Powerful, onset inhibition in the ventral nucleus of the lateral lemniscus. J Neurophysiol 94: 1651–1654, 2005.
Nayagam DA, Clarey JC, Paolini AG. Intracellular responses and morphology of rat ventral complex of the lateral lemniscus neurons in vivo. J Comp Neurol 498: 295–315, 2006.[CrossRef][ISI][Medline]
Nelson PG, Erulkar SD. Synaptic mechanisms of excitation and inhibition in the central auditory pathway. J Neurophysiol 26: 908–923, 1963.
N'Gouemo P, Morad M. Voltage-gated calcium channels in adult rat inferior colliculus neurons. Neuroscience 120: 815–826, 2003.[CrossRef][ISI][Medline]
O'Neill WE, Brimijoin WO. Directional selectivity for FM sweeps in the suprageniculate nucleus of the mustached bat medial geniculate body. J Neurophysiol 88: 172–187, 2002.
Park TJ, Pollak GD. GABA shapes sensitivity to interaural intensity disparities in the mustached bat's inferior colliculus: implications for encoding sound location. J Neurosci 13: 2050–2067, 1993.[Abstract]
Park TJ, Pollak GD. Azimuthal receptive fields are shaped by GABAergic inhibition in the inferior colliculus of the mustached bat. J Neurophysiol 72: 1080–1102, 1994.
Phillips DE, Irvine DRF. Responses of single units in physiologically defined primary auditory cortex (A1) of the cat frequency tuning and response to intensity, J Neurophysiol 45: 48–58, 1981.
Pollak GD, Park TJ. The effects of GABAergic inhibition on monaural response properties of neurons in the mustached bat's inferior colliculus. Hear Res 65: 99–117, 1993.[CrossRef][ISI][Medline]
Poon PW, Chen X, Cheung YM. Differences in FM response correlate with morphology of neurons in the rat inferior colliculus. Exp Brain Res 91: 94–104, 1992.[ISI][Medline]
Portfors CV, Felix RA. Spectral integration in the inferior colliculus of the CBA/CaJ mouse. Neuroscience 136: 1159–1170, 2005.[CrossRef][ISI][Medline]
Portfors CV, Wenstrup JJ. Responses to combinations of tones in the nuclei of the lateral lemniscus. J Assoc Res Otolaryngol 2: 104–117, 2001.[Medline]
Priebe NJ, Ferster D. Direction selectivity of excitation and inh
