Differences in FM response correlate with morphology of neurons in the rat inferior colliculus

Strong energy component is more important than spectral selectivity in modeling responses of midbrain auditory neurons to wide-band environmental sounds

Modeling central auditory neurons in response to complex sounds not only helps understanding neural processing of speech signals but can also provide insights for biomimetics in neuro-engineering. While modeling responses of midbrain auditory neurons to synthetic tones is rather good, modeling those to environmental sounds is less satisfactory. Environmental sounds typically contain a wide range of frequency components, often with strong and transient energy. These stimulus features have not been examined in the conventional approach of auditory modeling centered on spectral selectivity. To this end, we firstly compared responses to an environmental sound of auditory midbrain neurons across 3 subpopulations of neurons with frequency selectivity in the low, middle and high ranges; secondly, we manipulated the sound energy, both in power and in spectrum, and compared across these subpopulations how their modeled responses were affected. The environmental sound was recorded when a rat was drinking from a feeding bottle (called the 'drinking sound'). The sound spectrum was divided into 20 non-overlapping frequency bands (from 0 to 20 kHz, at 1 kHz width) and presented to an artificial neural model built on a committee machine with parallel spectral inputs to simulate the known tonotopic organization of the auditory system. The model was trained to predict empirical response probability profiles of neurons to the repeated sounds. Results showed that model performance depended more on the strong energy components than on the spectral selectivity. Findings were interpreted to reflect general sensitivity to rapidly changing sound intensities at the auditory midbrain and in the cortex.

Responses of midbrain auditory neurons to two different environmental sounds-A new approach on cross-sound modeling

When modeling auditory responses to environmental sounds, results are satisfactory if both training and testing are restricted to datasets of one type of sound. To predict 'cross-sound' responses (i.e., to predict the response to one type of sound e.g., rat Eating sound, after training with another type of sound e.g., rat Drinking sound), performance is typically poor. Here we implemented a novel approach to improve such cross-sound modeling (single unit datasets were collected at the auditory midbrain of anesthetized rats). The method had two key features: (a) population responses (e.g., average of 32 units) instead of responses of individual units were analyzed; and (b) the long sound segment was first divided into short segments (single sound-bouts), their similarity was then computed over a new metric involving the response (called Stimulus Response Model map or SRM map), and finally similar sound-bouts (regardless of sound type) and their associated responses (peri-stimulus time histograms, PSTHs) were modelled. Specifically, a committee machine model (artificial neural networks with 20 stratified spectral inputs) was trained with datasets from one sound type before predicting PSTH responses to another sound type. Model performance was markedly improved up to 92%. Results also suggested the involvement of different neural mechanisms in generating the early and late responses to amplitude transients in the broad-band environmental sounds. We concluded that it is possible to perform rather satisfactory cross-sound modeling on datasets grouped together based on their similarities in terms of the new metric of SRM map.

Functional organization of the local circuit in the inferior colliculus

The inferior colliculus (IC) is the first integration center of the auditory system. After the transformation of sound to neural signals in the cochlea, the signals are analyzed by brainstem auditory nuclei that, in turn, transmit this information to the IC. However, the neural circuitry that underlies this integration is unclear. This review consists of two parts: one is about the cell type which is likely to integrate sound information, and the other is about a technique which is useful for studying local circuitry. Large GABAergic (LG) neurons receive dense excitatory axosomatic terminals that originate from the lower brainstem auditory nuclei as well as local IC neurons. Dozens of axons coming from both local and lower brainstem neurons converge on a single LG soma. Excitatory neurons in IC can innervate many nearby LG somata in the same fibrodendritic lamina. The combination of local and ascending inputs is well suited for auditory integration. LG neurons are one of the main sources of inhibition in the medial geniculate body (MGB). LG neurons and the tectothalamic inhibitory system are present in a wide variety of mammalian species. This suggests that the circuitry of excitatory and inhibitory tectothalamic projections may have evolved earlier than GABAergic interneurons in the MGB, which are found in fewer species. Cellular-level functional imaging provides both morphological and functional information about local circuitry. In the last part of this review, we describe an in vivo calcium imaging study that sheds light on the functional organization of the IC.

Balance or imbalance: inhibitory circuits for direction selectivity in the auditory system

The auditory system detects and processes dynamic sound information transmitted in the environment. Other than the basic acoustic parameters, such as frequency, amplitude and phase, the time-varying changes of these parameters must also be encoded in our brain. Frequency-modulated (FM) sound is socially and environmentally significant, and the direction of FM sweeps is essential for animal communication and human speech. Many auditory neurons selectively respond to the directional change of such FM signals. In the past half century, our knowledge of auditory representation and processing has been updated frequently, due to technological advancement. Recently, in vivo whole-cell voltage clamp recordings have been applied to different brain regions in sensory systems. These recordings illustrate the synaptic mechanisms underlying basic sensory information processing and provide profound insights toward our understanding of neural circuits for complex signal analysis. In this review, we summarize the major findings of direction selectivity at several key auditory regions and emphasize on the recent discoveries on the synaptic mechanisms for direction selectivity in the auditory system. We conclude this review by describing promising technical developments in dissecting neural circuits and future directions in the study of complex sound analysis.

Modeling complex responses of FM-sensitive cells in the auditory midbrain using a committee machine

Frequency modulation (FM) is an important building block of complex sounds that include speech signals. Exploring the neural mechanisms of FM coding with computer modeling could help understand how speech sounds are processed in the brain. Here, we modeled the single unit responses of auditory neurons recorded from the midbrain of anesthetized rats. These neurons displayed spectral temporal receptive fields (STRFs) that had multiple-trigger features, and were more complex than those with single-trigger features. Their responses have not been modeled satisfactorily with simple artificial neural networks, unlike neurons with simple-trigger features. To improve model performance, here we tested an approach with the committee machine. For a given neuron, the peri-stimulus time histogram (PSTH) was first generated in response to a repeated random FM tone, and peaks in the PSTH were segregated into groups based on the similarity of their pre-spike FM trigger features. Each group was then modeled using an artificial neural network with simple architecture, and, when necessary, by increasing the number of neurons in the hidden layer. After initial training, the artificial neural networks with their optimized weighting coefficients were pooled into a committee machine for training. Finally, the model performance was tested by prediction of the response of the same cell to a novel FM tone. The results showed improvement over simple artificial neural networks, supporting that trigger-feature-based modeling can be extended to cells with complex responses. This article is part of a Special Issue entitled Neural Coding 2012. This article is part of a Special Issue entitled Neural Coding 2012.

Intracellular responses to frequency modulated tones in the dorsal cortex of the mouse inferior colliculus

Frequency modulations occur in many natural sounds, including vocalizations. The neuronal response to frequency modulated (FM) stimuli has been studied extensively in different brain areas, with an emphasis on the auditory cortex and the central nucleus of the inferior colliculus. Here, we measured the responses to FM sweeps in whole-cell recordings from neurons in the dorsal cortex of the mouse inferior colliculus. Both up- and downward logarithmic FM sweeps were presented at two different speeds to both the ipsi- and the contralateral ear. Based on the number of action potentials that were fired, between 10 and 24% of cells were selective for rate or direction of the FM sweeps. A somewhat lower percentage of cells, 6–21%, showed selectivity based on EPSP size. To study the mechanisms underlying the generation of FM selectivity, we compared FM responses with responses to simple tones in the same cells. We found that if pairs of neurons responded in a similar way to simple tones, they generally also responded in a similar way to FM sweeps. Further evidence that FM selectivity can be generated within the dorsal cortex was obtained by reconstructing FM sweeps from the response to simple tones using three different models. In about half of the direction selective neurons the selectivity was generated by spectrally asymmetric synaptic inhibition. In addition, evidence for direction selectivity based on the timing of excitatory responses was also obtained in some cells. No clear evidence for the local generation of rate selectivity was obtained. We conclude that FM direction selectivity can be generated within the dorsal cortex of the mouse inferior colliculus by multiple mechanisms.

The generation of direction selectivity in the auditory system

Both human speech and animal vocal signals contain frequency-modulated (FM) sounds. Although central auditory neurons that selectively respond to the direction of frequency modulation are known, the synaptic mechanisms underlying the generation of direction selectivity (DS) remain elusive. Here we show the emergence of DS neurons in the inferior colliculus by mapping the three major subcortical auditory nuclei. Cell-attached recordings reveal a highly reliable and precise firing of DS neurons to FM sweeps in a preferred direction. By using in vivo whole-cell current-clamp and voltage-clamp recordings, we found that the synaptic inputs to DS neurons are not direction selective, but temporally reversed excitatory and inhibitory synaptic inputs are evoked in response to opposing directions of FM sweeps. The construction of such temporal asymmetry, resulting DS, and its topography can be attributed to the spectral disparity of the excitatory and the inhibitory synaptic tonal receptive fields.

Modeling frequency modulated responses of midbrain auditory neurons based on trigger features and artificial neural networks

Frequency modulation (FM) is an important building block of communication signals for animals and human. Attempts to predict the response of central neurons to FM sounds have not been very successful, though achieving successful results could bring insights regarding the underlying neural mechanisms. Here we proposed a new method to predict responses of FM-sensitive neurons in the auditory midbrain. First we recorded single unit responses in anesthetized rats using a random FM tone to construct their spectro-temporal receptive fields (STRFs). Training of neurons in the artificial neural network to respond to a second random FM tone was based on the temporal information derived from the STRF. Specifically, the time window covered by the presumed trigger feature and its delay time to spike occurrence were used to train a finite impulse response neural network (FIRNN) to respond to this random FM. Finally we tested the model performance in predicting the response to another similar FM stimuli (third random FM tone). We found good performance in predicting the time of responses if not also the response magnitudes. Furthermore, the weighting function of the FIRNN showed temporal 'bumps' suggesting temporal integration of synaptic inputs from different frequency laminae. This article is part of a Special Issue entitled: Neural Coding.

Differential roles of GABAergic and glycinergic input on FM selectivity in the inferior colliculus of the pallid bat

Multiple mechanisms have been shown to shape frequency-modulated (FM) selectivity within the central nucleus of the inferior colliculus (IC) in the pallid bat. In this study we focus on the mechanisms associated with sideband inhibition. The relative arrival time of inhibition compared with excitation can be used to predict FM responses as measured with a two-tone inhibition paradigm. An early-arriving low-frequency inhibition (LFI) prevents responses to upward sweeps and thus shapes direction selectivity. A late-arriving high-frequency inhibition (HFI) suppresses slow FM sweeps and thus shapes rate selectivity for downward sweeps. Iontophoretic application of gabazine (GBZ) to block GABA(A) receptors or strychnine (Strych) to block glycine receptors was used to assess the effects of removal of inhibition on each form of FM selectivity. GBZ and Strych had a similar effect on FM direction selectivity, reducing selectivity in up to 86% of neurons when both drugs were coapplied. FM rate selectivity was more resistant to drug application with less than 38% of neurons affected. In addition, only Strych could eliminate FM rate selectivity, whereas GBZ alone was ineffective. The loss of FM selectivity was directly correlated to a loss of the respective inhibitory sideband that shapes that form of selectivity. The elimination of LFI correlated to a loss of FM direction selectivity, whereas elimination of HFI correlated to a loss of FM rate selectivity. Results indicate that 1) although the majority of FM direction selectivity is created within the IC, the majority of rate selectivity is inherited from lower levels of the auditory system, 2) a loss of LFI corresponds to a loss of FM direction selectivity and is created through either GABAergic or glycinergic input, and 3) a loss of HFI corresponds to a loss of FM rate selectivity and is created mainly through glycinergic input.

Should spikes be treated with equal weightings in the generation of spectro-temporal receptive fields?

Knowledge on the trigger features of central auditory neurons is important in the understanding of speech processing. Spectro-temporal receptive fields (STRFs) obtained using random stimuli and spike-triggered averaging allow visualization of trigger features which often appear blurry in the time-versus-frequency plot. For a clearer visualization we have previously developed a dejittering algorithm to sharpen trigger features in the STRF of FM-sensitive cells. Here we extended this algorithm to segregate spikes, based on their dejitter values, into two groups: normal and outlying, and to construct their STRF separately. We found that while the STRF of the normal jitter group resembled full trigger feature in the original STRF, those of the outlying jitter group resembled a different or partial trigger feature. This algorithm allowed the extraction of other weaker trigger features. Due to the presence of different trigger features in a given cell, we proposed that in the generation of STRF, the evoked spikes should not be treated indiscriminately with equal weightings.

Functional groups in the avian auditory system

Auditory perception depends on the coding and organization of the information-bearing acoustic features of sounds by auditory neurons. We report here that auditory neurons can be classified into functional groups, each of which plays a specific role in extracting distinct complex sound features. We recorded the electrophysiological responses of single auditory neurons in the songbird midbrain and forebrain to conspecific song, measured their tuning by calculating spectrotemporal receptive fields (STRFs), and classified them using multiple cluster analysis methods. Based on STRF shape, cells clustered into functional groups that divided the space of acoustical features into regions that represent cues for the fundamental acoustic percepts of pitch, timbre, and rhythm. Four major groups were found in the midbrain, and five major groups were found in the forebrain. Comparing STRFs in midbrain and forebrain neurons suggested that both inheritance and emergence of tuning properties occur as information ascends the auditory processing stream.

Intracellular recording reveals temporal integration in inferior colliculus neurons of awake bats

The central nucleus of the inferior colliculus (IC) is a major integrative center in the central auditory system. It receives information from both the ascending and descending auditory pathways. To determine how single IC neurons integrate information over a wide range of sound frequencies and sound levels, we examined their intracellular responses to frequency-modulated (FM) sounds in awake little brown bats (Myotis lucifugus). Postsynaptic potentials were recorded in response to downward FM sweeps of the range typical for little brown bats (80-20 kHz) and to three FM subcomponents (80-60, 60-40, and 40-20 kHz). The majority of recorded neurons responded to the 80- to 20-kHz downward FM sweep with a complex response. In this response an initial hyperpolarization was followed by depolarization with or without spike followed by hyperpolarization. Intracellular recordings in response to three FM subcomponents revealed that these neurons receive excitatory and inhibitory inputs from a wide range of sound frequencies. One third of IC neurons performed nearly linear temporal summation across a wide range of sound frequencies, whereas two thirds of IC neurons exhibited nonlinear summation with different degrees of nonlinearity. Some IC neurons showed different latencies of postsynaptic potentials in response to different FM subcomponents. Often responses to the later FM subcomponent occurred before responses to the earlier ones. This phenomenon may be responsible for response selectivity of IC neurons to FM sweeps.

FM signals produce robust paradoxical latency shifts in the bat's inferior colliculus

Previous studies in echolocating bats, Myotis lucifugus, showed that paradoxical latency shift (PLS) is essential for neural computation of target range and that a number of neurons in the inferior colliculus (IC) exhibit unit-specific PLS (characterized by longer first-spike latency at higher sound levels) in response to tone pulses at the unit's best frequency. The present study investigated whether or not frequency-modulated (FM) pulses that mimic the bat's echolocation sonar signals were equally effective in eliciting PLS. For two-thirds of PLS neurons in the IC, both FM and tone pulses could elicit PLS, but only FM pulses consistently produced unit-specific PLS. For the remainder of PLS neurons, only FM pulses effectively elicited PLS; these cells showed either no PLS or no response, to tone pulses. PLS neurons generally showed more pronounced PLS in response to narrow-band FM (each sweeping 20 kHz in 2 ms) pulse that contained the unit's best frequency. In addition, almost all PLS neurons showed duration-independent PLS to FM pulses, but the same units exhibited duration-dependent PLS to tone pulses. Taken together, when compared to tone pulses, FM stimuli can provide more reliable estimates of target range.

On the prediction of sweep rate and directional selectivity for FM sounds from two-tone interactions in the inferior colliculus

Two-tone stimuli have traditionally been used to reveal regions of inhibition in auditory spectral receptive fields, particularly for neurons with low spontaneous rates. These techniques reveal how different frequencies excite or suppress the response to an excitatory frequency of a cell, but have often been assessed at a fixed masker-probe time interval. We used a variation of this methodology to determine whether two-tone spectrotemporal interactions can account for rate-dependent directional selectivity for frequency modulations (FM) in the mustached bat inferior colliculus (IC). First, we quantified the response to upward and downward sweeping, linear, fixed-bandwidth FM tones centered at a unit's characteristic frequency (CF) at 6 sweep durations ranging from 2 to 64 ms. Then, to examine how responses to instantaneous frequencies contained within the sweeps might interact in time, we varied the frequency and relative onset of a brief (4 ms) "conditioner" tone paired with a fixed 4-ms CF probe tone. We constructed "conditioned response areas" (CRA) depicting regions of suppression and facilitation of the probe tone caused by the conditioning tone. We classified the CRAs as predominantly excitatory (40.9%), inhibitory (22.7%), or mixed (36.4%). To generate FM response predictions, the CRAs were multiplied with spectrograms of the same sweeps used to assess response to FM. The predictions of FM rate and directionality were accurate by our criteria in approximately 20% of units. Conversely, the CRAs from the remaining units failed to predict FM responses as accurately, suggesting that most IC units respond differently to FM sweeps than they do to tone-pairs matched to the instantaneous frequencies contained in those sweeps. The implications of these results for models of FM directionality are discussed.

Spectral integration in the inferior colliculus of the CBA/CaJ mouse

The inferior colliculus receives a massive convergence of inputs and in the mustached bat, this convergence leads to the creation of neurons in the inferior colliculus that integrate information across multiple frequency bands. These neurons are tuned to multiple frequency bands or are combination-sensitive; responding best to the combination of two signals of different frequency composition. The importance of combination-sensitive neurons in processing echolocation signals is well described, and it has been thought that combination sensitivity is a neural specialization for echolocation behaviors. Combination sensitivity and other response properties indicative of spectral integration have not been thoroughly examined in the inferior colliculus of non-echolocating mammals. In this study we tested the hypothesis that integration across frequencies occurs in the inferior colliculus of mice. We tested excitatory frequency response areas in the inferior colliculus of unanesthetized mice by varying the frequency of a single tone between 6 and 100 kHz. We then tested combination-sensitive responses by holding one tone at the unit's best frequency, and varying the frequency and intensity of a second tone. Thirty-two percent of the neurons were tuned to multiple frequency bands, 16% showed combination-sensitive facilitation and another 12% showed combination-sensitive inhibition. These findings suggests that the neural mechanisms underlying processing of complex sounds in the inferior colliculus share some common features among mammals as different as the bat and the mouse.

Processing of modulated sounds in the zebra finch auditory midbrain: responses to noise, frequency sweeps, and sinusoidal amplitude modulations

The avian auditory midbrain nucleus, the mesencephalicus lateralis, dorsalis (MLd), is the first auditory processing stage in which multiple parallel inputs converge, and it provides the input to the auditory thalamus. We studied the responses of single MLd neurons to four types of modulated sounds: 1) white noise; 2) band-limited noise; 3) frequency modulated (FM) sweeps, and 4) sinusoidally amplitude-modulated tones (SAM) in adult male zebra finches. Responses were compared with the responses of the same neurons to pure tones in terms of temporal response patterns, thresholds, characteristic frequencies, frequency tuning bandwidths, tuning sharpness, and spike rate/intensity relationships. Most neurons responded well to noise. More than one-half of the neurons responded selectively to particular portions of the noise, suggesting that, unlike forebrain neurons, many MLd neurons can encode specific acoustic components of highly modulated sounds such as noise. Selectivity for FM sweep direction was found in only 13% of cells that responded to sweeps. Those cells also showed asymmetric tuning curves, suggesting that asymmetric inhibition plays a role in FM directional selectivity. Responses to SAM showed that MLd neurons code temporal modulation rates using both spike rate and synchronization. Nearly all cells showed low-pass or band-pass filtering properties for SAM. Best modulation frequencies matched the temporal modulations in zebra finch song. Results suggest that auditory midbrain neurons are well suited for encoding a wide range of complex sounds with a high degree of temporal accuracy rather than selectively responding to only some sounds.

A new method for adjusting neural response jitter in the STRF obtained by spike-trigger averaging

Sensitivity of central auditory neurons to frequency modulated (FM) sound is often characterized based on spectro-temporal receptive field (STRF), which is generated by spike-trigger averaging a random stimulus. Due to the inherent property of time variability in neural response, this method erroneously represents the response jitter as stimulus jitter in the STRF. To reveal the trigger features more clearly, we have implemented a method that minimizes this error. Neural spikes from the brainstem of urethane-anesthetized rats were first recorded in response to two sets of FM stimuli: (a) a random FM tone for the generation of STRF and (b) a family of linear FM ramps for the determination of FM 'trigger point'. Based on the first dataset, STRFs were generated using spike-trigger averaging. Individual modulating waveforms were then matched with respect to their mean waveform at time-windows of a systematically varied length. A stable or optimal variance time profile was found at a particular window length. At this optimal window length, we performed delay adjustments. A marked sharpening in the FM bands in the STRF was found. Results were consistent with the FM 'trigger point' as estimated by the linear FM ramps. We concluded that the present approach of adjusting response jitter was effective in delineating FM trigger features in the STRF.

The effects of aging in the medial geniculate nucleus: a comparison with the inferior colliculus and auditory cortex

A common problem among the elderly is a difficulty in discriminating speech. One factor that may contribute to this is deterioration in the ability to process the dynamic components of speech such as formant transitions. The frequency-modulated (FM) sweep is a useful stimulus for investigating the neural basis of temporal processing speed since it has features in common with formant transitions. Previously, we showed that when cells in the auditory cortex of aged animals were presented with FM sweeps, they exhibited a decrease in temporal processing speed when compared to cells recorded from young animals. However, this was not the case for cells in the inferior colliculus (IC) where neural responses did not appear to be affected by aging. One question that remains is how the auditory thalamus is affected by aging: Is it similar to that of the auditory cortex or of the IC. To this end, single units were recorded from the ventral division of the medial geniculate nucleus (MGNv) of young and aged anaesthetized rats in response to FM sweeps. Results showed that there were no age-related differences in speed or direction selectivity of FM sweep responses in the MGNv. When compared with units recorded from the IC and AI, the responses of MGNv neurons were similar to those of the IC. This suggests that temporal processing speed is affected by aging in the cortex, but not in the auditory thalamus or midbrain.

Frequency modulated sweep responses in the medial geniculate nucleus

A basic feature of communication signals is a dynamic change in frequency. One stimulus that lends itself well to investigating the frequency changes contained in these signals is the frequency modulated (FM) sweep. While many studies have investigated FM sweep responses in the auditory midbrain and cortex, relatively few have examined them in the thalamus. To this end, we investigated the responses of single units in the ventral division of the medial geniculate nucleus (MGNv) of the rat to FM sweeps. Both upward- (changing from low to high frequency) and downward-directed (changing from high to low frequency) FM sweeps were presented at four rates of frequency modulation (i.e., speed). Results showed that the majority (76%) of the cells preferred fast or medium FM sweeps. For direction selectivity, just under half of the units (47%) exhibited a preference for the direction of FM sweep. The results suggest that there is a greater degree of direction but not speed selectivity at progressively higher levels in the auditory pathway.

Expression of the Kv1.1 ion channel subunit in the auditory brainstem of the big brown bat, Eptesicus fuscus

Voltage-gated potassium channels play an important role in shaping membrane properties that underlie neurons' discharge patterns and the ways in which they transform their input. In the auditory system, low threshold potassium currents such as those created by Kv1.1 subunits contribute to precise phaselocking and to transient onset responses that provide time markers for temporal features of sounds. The purpose of the present study was to compare information about the distribution of neurons expressing the KV 1.1 in the brainstem auditory nuclei with the distribution of neurons with known functional properties in the auditory system of the big brown bat, Eptesicus fuscus. We used immunocytochemistry and light microscopy to look at the distribution of Kv1.1 subunits in the brainstem auditory nuclei. There was prominent expression in cell types known to contain high levels of Kv1.1 in other species and known to respond to auditory signals with high temporal precision. These included octopus cells and spherical bushy cells of the cochlear nucleus and principal neurons of the medial nucleus of the trapezoid body. In addition, we found high levels of Kv1.1 in neurons of the columnar subdivision of the ventral nucleus of the lateral lemniscus and in ventral periolivary cell groups. Neurons with high levels of Kv1.1 were differentially distributed in the intermediate nucleus of the lateral lemniscus and in the inferior colliculus, suggesting that these structures contain functionally distinct cell populations, some of which may be involved in high-precision temporal processing.

Temporal processing speed in the inferior colliculus of young and aged rats

A common problem among the elderly is a difficulty in discriminating speech. One factor that may contribute to this is a deterioration in the ability to process dynamic aspects of speech such as formant transitions. Recently, Mendelson and Ricketts [Mendelson, J.R., Ricketts, C., Hear. Res. 158 (2001) 84-94] showed that cells recorded from the auditory cortex of aged animals exhibited a decrease in temporal processing speed compared to young animals. In the present study, we examined whether this age-related effect was exclusive to the auditory cortex or whether it was apparent subcortically. To this end, single units were recorded from the inferior colliculus (IC) of young and aged rats in response to frequency modulated (FM) sweeps. Results showed that there was no age-related difference in speed or direction selectivity of FM sweep responses in the IC. The present results suggest that the effect of aging on temporal processing speed occurs in the cortex, but not subcortically.

Between sound and perception: reviewing the search for a neural code

This review investigates the roles of representation, transformation and coding as part of a hierarchical process between sound and perception. This is followed by a survey of how speech sounds and elements thereof are represented in the activity patterns along the auditory pathway. Then the evidence for a place representation of texture features of sound, comprising frequency, periodicity pitch, harmonicity in vowels, and direction and speed of frequency modulation, and for a temporal and synchrony representation of sound contours, comprising onsets, offsets, voice onset time, and low rate amplitude modulation, in auditory cortex is reviewed. Contours mark changes and transitions in sound and auditory cortex appears particularly sensitive to these dynamic aspects of sound. Texture determines which neurons, both cortical and subcortical, are activated by the sound whereas the contours modulate the activity of those neurons. Because contours are temporally represented in the majority of neurons activated by the texture aspects of sound, each of these neurons is part of an ensemble formed by the combination of contour and texture sensitivity. A multiplexed coding of complex sound is proposed whereby the contours set up widespread synchrony across those neurons in all auditory cortical areas that are activated by the texture of sound.

Responses of young and aged rat inferior colliculus neurons to sinusoidally amplitude modulated stimuli

The inferior colliculus (IC) is a processing center for monaural and binaural auditory signals. Many units in the central nucleus of the inferior colliculus (CIC) respond to amplitude and frequency modulated tones, features found in communication signals. The present study examined potential effects of age on responses to sinusoidally amplitude modulated (SAM) tones in CIC and external cortex of the inferior colliculus (ECIC) units in young and aged F344 rats. Extracellular recordings from 154 localized single units of aged (24 month) rats were compared to recordings from 135 IC units from young adult (3 month) animals. SAM tones were presented at 30 dB above threshold. Comparisons were made between CIC and ECIC regarding the percentage of units responding to SAM stimuli, the relationship between SAM responsiveness and temporal response patterns, maximum discharge rates and maximum modulation gains, shapes of rate transfer functions and synchronization modulation transfer functions (MTFs) in response to SAM tones. Sixty percent of units in young and aged rat IC were selectively responsive to SAM stimuli. Eighty-one percent of units classified as onset temporal response patterns were not tonically responsive to SAM stimuli. Median maximum discharge rate in response to SAM tones was 17.6/s in young F344 rats; median maximum modulation gain was 3.85 dB. These measurements did not change significantly with age. Thirty-seven percent of young rat units displayed bandpass MTFs and 53% had lowpass MTFs. There was a significant age-related shift in the distribution of MTF shapes in both the CIC and ECIC. Aged animals showed a lower percentage of bandpass functions and a higher percentage of lowpass functions. Age-related changes observed in SAM coding may reflect an altered balance between excitatory/inhibitory neurotransmitter efficacy in the aged rat IC, and/or possibly a change in the functional dynamic range of IC neurons.

Similarities of FM and AM receptive space of single units at the auditory midbrain

Complex sounds, including human speech, contain time-varying signals like frequency modulation (FM) and amplitude modulation (AM) components. In spite of various attempts to characterize their neuronal coding in the mammalian auditory systems, a unified view of their responses has not been reached. We compared FM and AM coding in terms of receptive space with reference to the input-output relationship of the underlying neural circuits. Using extracellular recording, single unit responses to a novel stimulus (i.e. random AM or FM tone) were obtained at the auditory midbrain of the anesthetized rat. Responses could be classified into three general types, corresponding to selective sensitivity to one of the following aspects of the modulation: (a) steady state, (b) dynamic state, or (c) steady-and-dynamic states. Such response typing was basically similar between FM and AM stimuli. Furthermore, the receptive space of each unit could be characterized in a three-dimensional Cartesian co-ordinate system formed by three modulation parameters: velocity, range and intensity. This representation applies to both FM and AM responses. We concluded that the FM and AM codings are very similar at the auditory midbrain and may likely involve similar neural mechanisms.

Neural measurement of sound duration: control by excitatory-inhibitory interactions in the inferior colliculus

In the inferior colliculus (IC) of the big brown bat, a subpopulation of cells ( approximately 35%) are tuned to a narrow range of sound durations. Band-pass tuning for sound duration has not been seen at lower levels of the auditory pathway. Previous work suggests that it arises at the IC through the interaction of sound-evoked, temporally offset, excitatory and inhibitory inputs. To test this hypothesis, we recorded from duration-tuned neurons in the IC and examined duration tuning before and after iontophoretic infusion of antagonists to gamma-aminobutyric acid-A (GABA(A)) (bicuculline) or glycine (strychnine). The criterion for duration tuning was that the neuron's spike count as a function of duration had a peak value at one duration or a range of durations that was >/=2 times the lowest nonzero value at longer durations. Out of 21 units tested with bicuculline, duration tuning was eliminated in 15, broadened in two, and unaltered in four. Out of 10 units tested with strychnine, duration tuning was eliminated in four, broadened in one, and unaltered in five. For units tested with both bicuculline and strychnine, bicuculline had a greater effect on reducing or abolishing duration tuning than did strychnine. Bicuculline and strychnine both produced changes in discharge pattern. There was nearly always a shift from an offset response to an onset response, indicating that in the predrug condition, inhibition arrived simultaneously with excitation or preceded it. There was often an increase in the length of the spike train, indicating that in the predrug condition, inhibition also coincided with later parts of excitation. These findings support two hypotheses. First, duration tuning is created in the IC. Second, although the construction of duration tuning varies in some details among IC neurons, it follows three rules: 1) an excitatory and an inhibitory event are temporally linked to the onset of sound but temporally offset from one another; 2) the duration of some inhibitory event must be linked to the duration of the sound; 3) an excitatory event must be linked to the offset of sound.

Processing of sinusoidally frequency modulated signals in the nuclei of the lateral lemniscus of the big brown bat, Eptesicus fuscus

Neurons in the nuclei of the lateral lemniscus (NLL) of the big brown bat, Eptesicus fuscus, show several distinctive patterns of response to unmodulated tones. Previous work suggests that sustained responders are specialized to transmit information about sound level and duration while onset responders transmit precise timing information. The biosonar signals of E. fuscus consist of multiple, downward frequency modulated sweeps that change in slope and repetition rate as the bat approaches a target. An obvious hypothesis would be that NLL neurons with sustained responses should discharge during the time when the frequency of a signal is within their response area, but that onset responders should discharge each time the frequency enters the excitatory portion of their response area. In this study we examined the responses of NLL neurons to sinusoidally frequency modulated (SFM) signals presented monaurally to awake, restrained bats. Extracellular recordings were obtained from single neurons in the multipolar and columnar divisions of the ventral nucleus (VNLLm and VNLLc), the intermediate nucleus (INLL) and the dorsal nucleus of the lateral lemniscus (DNLL). All NLL neurons responded synchronously to SFM signals under some conditions. The temporal precision of synchronization was quantified using a coefficient of synchronization (CS), where a value of I equals perfect synchrony. Maximum CS values ranged from 0.70 to >0.99, were generally highest at low modulation rates ( <200 Hz), and showed lowpass characteristics for modulation rate. The maximal modulation rates that elicited synchronous discharge ranged from 50 to 500 Hz. The highest maximal rates were found in the VNLLm and VNLLc, the lowest in DNLL. The ability of NLL neurons to synchronize their discharge to the pattern of an SFM signal is intermediate between that of neurons in the cochlear nucleus and in the inferior colliculus. For the majority of neurons in VNLLm, INLL and DNLL, the precision of synchronization was approximately equal for the downward and upward components of the SFM signal; in contrast, 69% of VNLLc neurons responded selectively to the downward component of the SFM signal. All VNLLc neurons and a subset of those in VNLLm, INLL, and DNLL responded synchronously to SFM signals only if the frequency excursions included a border of the excitatory frequency bandwidth, suggesting that the synchronous discharge was due primarily to the repeated passage of the stimulus frequency into and out of the excitatory portion of the response area. In the case of VNLLc neurons, only the high frequency border was effective; Other neurons, especially those in DNLL, responded synchronously to SFM signals with frequency excursions that were confined entirely within the excitatory response area.

Temporal processing across frequency channels by FM selective auditory neurons can account for FM rate selectivity

Auditory neurons tuned to the direction and rate of frequency modulations (FM) might underlie the encoding of frequency sweeps in animal vocalizations and formant transitions in human speech. We examined the relationship between FM direction and rate selectivity and the precise temporal interactions of excitatory and inhibitory sideband inputs. Extracellular single-unit recordings were made in the auditory midbrains of eight mustached bats. Up- and down-sweeping linear FM stimuli were presented at different modulation rates in order to determine FM selectivity. Brief tone pairs with varying interstimulus delays were presented in a forward masking paradigm to examine the relative timing of excitatory and inhibitory inputs. In the 33 units for which tone pair data were collected, a correspondence existed between FM rate selectivity and the time delays between paired tones. Moreover, FM directional selectivity was strongly linked to rate selectivity, because directional preferences were expressed only at certain rates and not others. We discuss how abnormalities in the relative timing of inputs could alter or abolish the selectivity of such neurons, and how such a mechanism could account for the perceptual deficits for formant transitions seen in certain children with phonological deficits.

GABAergic and glycinergic inhibition sharpens tuning for frequency modulations in the inferior colliculus of the big brown bat

Discrimination of amplitude and frequency modulated sounds is an important task of auditory processing. Experiments have shown that tuning of neurons to sinusoidally frequency- and amplitude-modulated (SFM and SAM, respectively) sounds becomes successively narrower going from lower to higher auditory brain stem nuclei. In the inferior colliculus (IC), many neurons are sharply tuned to the modulation frequency of SFM sounds. The purpose of this study was to determine whether GABAergic or glycinergic inhibition is involved in shaping the tuning for the modulation frequency of SFM sounds in IC neurons of the big brown bat (Eptesicus fuscus). We recorded the response of 56 single units in the central nucleus of the IC to SFM stimuli before and during the application of the gamma-aminobutyric acid-A (GABAA) receptor antagonist bicuculline or the glycine receptor antagonist strychnine. To evaluate tuning to the modulation frequency, the normalized spike count (normalized according to the maximal response for each condition tested) was plotted versus the modulation frequency and the upper and lower 50% cutoff points were determined. Bicuculline increased the upper cutoff in 46% of the neurons by >/=25%. The lower cutoff decreased in 48% of the neurons tested. In some neurons (approximately 30%), a sharpening of the tuning by bicuculline was observed. Strychnine induced an increase of the upper cutoff in almost half of the neurons. Compared with bicuculline these changes were smaller. The lower cutoff decreased in 50% of the neurons with strychnine. The synchronization coefficient (SC) was calculated and compared for three modulation frequencies (50, 100, and 200 Hz) between predrug and drug condition. For all neurons, synchronization decreased (n = 36) or did not change (n = 26) during drug application. This was mainly an effect of the prolonged discharge in response to each cycle. Under predrug conditions, many neurons exhibited selectivity to the direction of the FM, hence they only responded once to each cycle. In a minority of neurons, direction selectivity was abolished by drug application. The main finding was that neuronal inhibition sharpens tuning to the modulation frequency in the majority of neurons. In general, changes induced by bicuculline or strychnine were comparable.

Modeling of the response of midbrain auditory neurons in the rat to their vocalization sounds based on FM sensitivities

Single units were recorded from the inferior colliculus (IC) of anaesthetized rats in response to: (a) an FM tone, the frequency of which was randomly varied, and (b) a digitized rat vocalization sound. We hypothesized that these neurons may have 'orientation-specific' spectrotemporal receptive field (STRF) that can be used to estimate their responses to complex communication signals. Based on the FM response, we first estimated the cell's STRF which was then convolved with the spectrogram of the rat's vocalization call. A simple convolution gave only crude prediction of the cell's response to the vocalization sound. When inhibitory areas were added around certain parts of the STRF, a better match was found. We conclude that for some FM-sensitive neurons of the IC, STRF with inhibitory areas may account for their responses to vocalization sounds.

Responses to exponential frequency modulations in the rat inferior colliculus

We examined responses to pure tones and exponentially frequency-modulated (FM) stimuli in the inferior colliculus of ketamine anesthetized rats. All units responded to both pure-tone and FM stimulation: units responding selectively to FM stimuli were not found. The comparison between responses to many different FM sweeps revealed that activity was elicited when the instantaneous frequency of a FM sweep entered the unit's pure-tone tuning curve. Units were tuned to the rate of frequency modulation. Most modulation rate transfer functions had bandpass characteristics. Best modulation rates covered a range from 4.8 to 1904 octaves/s with more than 90% between 10 and 400 octaves/s. In contrast to previous studies, modulation direction was not coded in unit responses and only few units demonstrated a weak change in response strength when sweep direction was altered. This is at least partly attributable to the FM stimulus design which, in the present study, was adapted to the logarithmic frequency representation in the rat auditory system and carefully matched to the units' pure-tone responses area. In spite of the close relationship between pure tone and FM response behavior, modulation rate tuning cannot be completely explained on the basis of the units' pure-tone responses.

Processing of modulation frequency in the dorsal cochlear nucleus of the guinea pig: sinusoidal frequency-modulated tones

Frequency- and amplitude-modulated (FM and AM, respectively) tones are important information-bearing elements in voice sounds and can also be produced by the spatial movement of sound sources. Zhao and Liang (1995) recently reported the response features of dorsal cochlear nucleus (DCN) neurons to AM tones. In the present study, the responses of the guinea pig DCN neurons to sinusoidal FM (SFM) tones were examined. Discharges of the DCN units to the SFM tones phase-locked to the stimulus modulation frequencies (fm). The phase-locked discharge patterns existed over broad ranges of modulation parameters and at stimulus levels as high as 95 dB SPL or modulation depths (dm) as low as 2%. Robust phase-locking to the fm was observed in samples of all DCN unit types studied. The means of best fm (Bfm) and upper limit fm (ULfm) of all recorded units were 510 Hz and 940 Hz, respectively. Pauser/Buildup (P/B) units had mean maximum synchronization index (SImax) of 0.57. ON units had the highest Bfm with the mean of 646 Hz and subtype ON-S showed the highest mean of SImax at 0.63. Phase-locking to the fm was independent of discharge rates and existed even when the discharge rates were reduced to the background spontaneous rate (SR). A few units showed stronger synchronous responses to the square and triangular FM stimuli instead of the SFM tones. The relationship between the modulated responses and the unit's response area were further examined. The fm phase-locking occurred to modulation bands (or frequency ranges) within the response area, with the modulation bands as narrow as +/- 160 Hz in the central inhibitory areas of the type IV units. As the width of the modulation band changed within a unit's response area, the phases of the fm phase-locked responses of P/B units linearly changed while for Onset units, the change was lesser. The P/B and Onset units had a pi phase shift and a pi/2 phase change, respectively, as carrier frequencies (fc5) passed through characteristic frequencies (CF) and the excitatory/inhibitory response boundaries. The phase-locked responses to the fms were dependent on the SR but were independent of the CF. Low-SR (< or = 2 spikes/s) units had higher synchronization of responses to the fm than the high-SR (> 2 spikes/s) units (SImax = 0.64 and 0.42 respectively). These results suggest that the temporal characteristics of the fm is effectively represented in the responses of DCN units to the SFM tones. Such temporal encoding behavior can play an important role in the processing of the complex sounds in the auditory system. These results also have implications for a possible role for the DCN is in identifying the spatial movement of a sound source.

Comparison of the topology and growth rules of motoneuronal dendrites

The complexity, shape, and branching modes of the dendrites of spinal motoneurons were compared in cat, rat, and frog using topological analysis and growth models. The complexity of motoneuronal dendrites, measured as the mean number of terminal segments, varied significantly among samples and was related to contractile properties of innervated motor units. Despite this variation, all mature motoneurons having a mean number of terminal segments per dendrite greater than ten (up to 24.3) exhibited a narrow range of values of coefficients describing the symmetry of tree shapes (0.42-0.47). This implies low variability in the topological shape of motoneuronal dendrites of different animals. This similarity of tree shapes proved to be a result of the similarity of growth rules. The growth of the dendrites could be described to a first approximation by a two-parameter (Q and S) model called the QS model and by a multitype Markovian model. The estimation of parameters of the QS model, in which parameter Q is related to the probability of branching of intermediate segments, revealed that Q was equal or close to 0, implying that branching of dendrites is restricted to terminal segments. The estimates of the parameter S, which describes whether the probability of branching increases (S < 0) or decreases (S > 0) exponentially with segment order, were positive. This was in agreement with the results of estimation of probabilities of branching provided by the Markovian model, which showed that the branching probabilities decreased with segment order in an exponential manner in most of the neurons studied. The QS and Markovian models involve different assumptions about the sequence and timing of branching events, and selection of the best model can provide insight into details of dendritic outgrowth. Extensive simulation of tree outgrowth using a Markovian model revealed significant differences between stimulated trees and real dendrites, particularly with regard to variability of the number of terminals and to symmetry. In contrast, the QS model provided a good fit to the mean values and standard deviations of basic topological parameters. This model is adequate to describe the shape of mature motoneuronal dendrites. It implies that dendritic branches have many opportunities to bifurcate during the whole time of development and that bifurcating potency of a branch is a function of the number and position of other branches of that dendrite. Combined with analysis of metrical properties such as lengths of segments, the QS model can assist in a quantitative analysis of development and plasticity.