Perception and selectivity of sound duration in the central auditory midbrain / 生理学报

Sound duration plays important role in acoustic communication. Information of acoustic signal is mainly encoded in the amplitude and frequency spectrum of different durations. Duration selective neurons exist in the central auditory system including inferior colliculus (IC) of frog, bat, mouse and chinchilla, etc., and they are important in signal recognition and feature detection. Two generally accepted models, which are "coincidence detector model" and "anti-coincidence detector model", have been raised to explain the mechanism of neural selective responses to sound durations based on the study of IC neurons in bats. Although they are different in details, they both emphasize the importance of synaptic integration of excitatory and inhibitory inputs, and are able to explain the responses of most duration-selective neurons. However, both of the hypotheses need to be improved since other sound parameters, such as spectral pattern, amplitude and repetition rate, could affect the duration selectivity of the neurons. The dynamic changes of sound parameters are believed to enable the animal to effectively perform recognition of behavior related acoustic signals. Under free field sound stimulation, we analyzed the neural responses in the IC and auditory cortex of mouse and bat to sounds with different duration, frequency and amplitude, using intracellular or extracellular recording techniques. Based on our work and previous studies, this article reviews the properties of duration selectivity in central auditory system and discusses the mechanisms of duration selectivity and the effect of other sound parameters on the duration coding of auditory neurons.

Effects of modulation range and presentation rate of FM stimulus on auditory response properties of mouse inferior collicular neurons / 生理学报

In natural acoustical environments, most biologically related sounds containing frequency-modulated (FM) components repeat over periods of time. They are often in rapid sequence rather than in temporal isolation. Few studies examined the neuronal response patterns evoked by FM stimuli at different presentation rates (PR). In the present investigation, by using normal electrophysiological technique, we specifically studied the temporal features of response of the inferior collicular (IC) neurons to FM sweeps with different modulation ranges (MR) in conditions of distinct PR in mouse. The results showed that most of the recorded neurons responded best to narrower MRs (narrow-pass, up-sweep: 60.00%, 54/90; down-sweep: 63.33%, 57/90), while a small fraction of neurons displayed other patterns such as band-pass (up-sweep, 16.67%, 15/90; down-sweep, 18.89%, 17/90), all-pass (up-sweep, 18.89%, 17/90; down-sweep, 13.33%, 12/90) and wide-pass (up-sweep, 4.44%, 4/90; down-sweep, 4.44%, 4/90). Both the discharge rate and duration of recorded neurons decreased but the latency lengthened with increase in PR, when different PRs from 0.5/s to 10/s of FM sound were used. The percentage of total directional selective neurons, up-directional selective neurons, and down-directional selective neurons changed with the variation of PR or MR. These results indicate that temporal features of mouse midbrain neurons responding to FM sweeps are co-shaped by the MR and PR. Possible mechanisms underlying may be related to spectral and temporal integration of the FM sound by the IC neurons.

Sound duration and sound pattern affect the recovery cycles of inferior collicular neurons in leaf-nosed bat, Hipposideros armiger / 生理学报

The effects of sound duration and sound pattern on the recovery cycles of inferior collicular (IC) neurons in constant frequency-frequency modulation (CF-FM) bats were explored in this study. Five leaf-nosed bats, Hipposideros armiger (4 males, 1 female, 43-50 g body weight), were used as subjects. The extracellular responses of IC neurons to paired sound stimuli with different duration and patterns were recorded, and the recovery was counted as the ratio of the second response to the first response. Totally, 169 sound-sensitive IC neurons were recorded in the experiment. According to the interpulse interval (IPI) of paired sounds when neurons reached 50% recovery (50% IPI), the recovery cycles of these IC neurons were classified into 3 types: fast recovery (F, the 50% IPI was less than 15 ms), short recovery (S, the 50% IPI was between 15.1 and 30 ms) and long recovery (L, the 50% IPI was more than 30 ms). When paired CF stimuli with 2 ms duration was used, the ratio of F neurons was 32.3%, and it decreased to 18.1% and 18.2% respectively when 5 and 7 ms CF stimuli were used. The ratios of S and L neurons were 41.5%, 33.7%, 29.1% and 26.2%, 48.2%, 52.7% respectively when 2, 5 and 7 ms CF stimuli were used. The average 50% IPI determined after stimulation with paired 2 ms, 5 ms and 7 ms CF sounds were (30.2 ± 27.6), (39.9 ± 29.1) and (49.4 ± 34.7) ms, respectively, and the difference among them was significant (P< 0.01). When the stimuli of paired 2 ms CF sounds were shifted to paired 2 ms FM sounds, the proportion of F, S and L neurons changed from 32.3%, 41.5%, 26.2% to 47.7%, 24.6%, 27.7%, respectively, and the average 50% IPI decreased from (30.2 ± 27.6) to (23.9 ± 19.0) ms (P< 0.05, n = 65). When paired 5+2 ms CF-FM pulses were used instead of 7 ms CF sounds, the proportion of F, S and L neurons changed from 18.2%, 29.1%, 52.7% to 29.1%, 27.3%, 43.6%, respectively, and the average 50% IPI decreased from (49.4 ± 34.7) to (36.3 ± 29.4) ms (P< 0.05, n = 55). All these results suggest that the CF and FM components in echolocation signal of CF-FM bats play different roles during bats' hunting and preying on. The FM component of CF-FM signal presenting in the terminal phase can increase the number of F type neurons and decrease the recovery cycles of IC neurons for processing high repetition echo information, which ensures the bat to analyze the target range and surface texture more accurately.

Selectivity for the rate of frequency-modulated sweeps and its affecting factors in the inferior collicular neurons of mouse / 生理学报

Both animal communication sounds and human speech contain frequency-modulated (FM) sweeps. Although the selectivity for the rate of FM sweeps in neurons has been found in many kinds of animals at different levels of the central auditory structures, the underlying neural mechanism is still not clear. Using extracellular single unit recording techniques, we examined the selectivity for the rate of FM sweeps in the inferior colliculus (IC) neurons of the Kunming mouse (Mus musculus, Km) in the free-field stimulation conditions and determined its affecting factors. Totally, 102 neurons were recorded successfully, among which 42 neurons (41.2%) displayed a duration tuning pattern under pure tone (PT) stimulus. The percentages of short-pass, band-pass, and long-pass neurons were 22.6% (23/10), 8.8% (9/102), 9.8% (10/102), respectively. The other 60 neurons (58.8%) did not show any duration tuning features. Under FM stimulus, the majority of duration tuning neurons (78.6%, 33/42) showed the selectivity for the rate of FM sweeps. For these neurons, the type of rate selectivity was determined by the duration tuning features, but it was not related to the modulation range (MR) of FM. In a small fraction of duration tuning neurons (21.4%, 9/42), the rate selectivity was correlated with the MR, but uncorrelated with the duration tuning features. On the other hand, more than half of the non-duration tuning neurons (53.3%, 32/60) exhibited the rate selectivity under FM stimulus, and almost all of them (31/32) showed fast-rate selectivity. Nevertheless, there were 8 neurons (in 32) displaying the same best rate at different MR, indicating that they were real rate-selective neurons. Our results indicate that the selectivity for the rate of FM sweeps is co-determined by duration tuning features and sweep bandwidth. Only a few of inferior colliculus neurons belong to real rate-selectivity neurons in the mouse.

GABAergic inhibition modulates intensity sensitivity of temporally patterned pulse trains in the inferior collicular neurons in big brown bats / 生理学报

The echolocating big brown bats (Eptesicus fuscus) emit trains of frequency-modulated (FM) biosonar signals with duration, amplitude, repetition rate, and sweep structure changing systematically during interception of their prey. In the present study, the sound stimuli of temporally patterned pulse trains at three different pulse repetition rates (PRRs) were used to mimic the sounds received during search, approach, and terminal stages of echolocation. Electrophysiological method was adopted in recordings from the inferior colliculus (IC) of midbrain. By means of iontophoretic application of bicuculline, the effect of GABAergic inhibition on the intensity sensitivity of IC neurons responding to three different PRRs of 10, 30 and 90 pulses per second (pps) was examined. The rate-intensity functions (RIFs) were acquired. The dynamic range (DR) of RIFs was considered as a criterion of intensity sensitivity. Comparing the average DR of RIFs at different PRRs, we found that the intensity sensitivity of some neurons improved, but that of other neurons decayed when repetition rate of stimulus trains increased from 10 to 30 and 90 pps. During application of bicuculline, the number of impulses responding to the different pulse trains increased under all stimulating conditions, while the DR differences of RIFs at different PRRs were abolished. The results indicate that GABAergic inhibition was involved in modulating the intensity sensitivity of IC neurons responding to pulse trains at different PRRs. Before and during bicuculline application, the percentage of changes in responses was maximal in lower stimulus intensity near to the minimum threshold (MT), and decreased gradually with the increment of stimulus intensity. This observation suggests that GABAergic inhibition contributes more effectively to the intensity sensitivity of the IC neurons responding to pulse trains at lower sound level.

Different forward masking patterns of sustained noise burst and segmental noise burst in the inferior collicular neurons of the mouse / 生理学报

Although there has been a growing body of literature showing the neural correlation of forward masking caused by a pure tone masker in the auditory neurons, relative few studies have addressed the description of how the forward masking caused by a noise burst, especially a sequence of noise burst, is transformed into neuronal representation in the central auditory system. Using a noise forward masking paradigm under free field stimuli conditions, this in vivo study was devoted to exploring it in the inferior collicular (IC) neurons of the mouse (Mus musculus KM). A total of 96 IC neurons were recorded. Rate-intensity functions (RIFs) with and without the presentation of masker, sustained noise burst (SNB) or segmental noise burst (SGNB), were measured in 51 neurons. We found that the relative masker intensities were distributed over a wide range between 21 dB below the minimum threshold (MT) and 19 dB above the MT of the corresponding probe tone. The masking effect of the SGNB on firing rate in nearly half of neurons (type I, 45.10%) was stronger than that of the SNB (P<0.001), whereas in a smaller fraction of neurons (type III, 17.65%), it was weaker than that of the SNB (P<0.001). There was no significant difference in masking effect between the SNB and SGNB in type II neurons (37.25%, P>0.05). Irrespective of type I or type III neurons, the inhibitory effects of both kinds of maskers were all greater at lower probe intensities but decreased significantly with the increase of probe intensity (P<0.001). Interestingly, as the probe intensity increased, the difference of masking effect between the SNB and SGNB disappeared (P>0.05). In addition, we observed that temporal masking pattern could be transformed when the masker was changed from the SNB to SGNB. The main type of this transformation was from early-inhibition to proportional-inhibition pattern (53.85%, 7/13). Our data provide the evidence that the inhibitory effects of these two maskers have differential weights over time and intensity domains of the IC neurons responding to a pure tone. This suggests that the forward masking of noise is by no means the source of simply suppression in neuronal firing rate. There might be a few of active neural modulating ways in which the coding of temporal acoustical information can be operated.

Effects of backward masking on the responses of the inferior collicular neurons in the big brown bat, Eptesicus fuscus / 生理学报

Temporal features of sound convey information vital for behaviors as diverse as speech recognition by human and echolocation by bats. However, auditory stimuli presented in temporal proximity might interfere with each other. Although much progress has been made in the description of this phenomenon from psychophysical studies, the neural mechanism responsible for its formation at central auditory structures especially at the inferior colliculus (IC), a midbrain auditory nucleus which practically receives massive bilateral projections from all the major auditory structures in the brainstem, remains unclear. This study was designed to investigate it in vivo by using electrophysiological recording from the inferior collicular neurons of the big brown bat, Eptesicus fuscus. In our results, the responses of 12 (38%, n= 31) neurons to the test sound (leading sound) were obviously inhibited by the masker (lagging sound). The inhibitory effects in these neurons were correlated with the inter-stimulus level difference (SLD) and the inter-stimulus onset asynchrony (SOA) interval. The strength of backward masking increased with the masker intensity increasing, the test sound intensity decreasing and the SOA interval shortening. There were no obvious effects of backward masking on the responses of many other neurons (52%, 16/31), and yet in a part of these neurons, the neural inhibition of responses to the test sound was observed at the special SLD and the special SOA intervals. Moreover, few of the 31 sampled IC neurons (10%, 3/31) displayed facilitating responses to the test sound at the special SLD and the special SOA intervals. These data demonstrate that a lot of IC neurons are involved in the generation of the backward masking of acoustical perception. It is conjectured that the temporal dynamic integration between the leading inhibitory inputs evoked by the masker sound and the excitatory inputs evoked by the test sound might play a key role in shaping the acoustical response characteristics of the IC neurons.

Dynamic modulations on intensity sensitivity evoked by weak noise in the inferior collicular neurons / 生理学报

In order to explore the possible mechanisms by which ethologically relevant sounds can be extracted from complex auditory environments, this study examined the effects of weak noise on the rate-intensity functions (RIFs) of neurons responding to tone burst in the inferior colliculus (IC) of nine mice (Mus musculus Km). Under free field stimuli conditions, a total of 112 IC neurons were recorded. RIFs with and without simultaneous presentation of weak noise, of which the intensity was relative to 5 dB below minimum threshold of tone burst, were measured in 44 IC neurons. By means of evaluating the changes of dynamic range (DR), slope of RIFs, and percent inhibition at different tone burst intensities evoked by the weak noise, three types of variations in RIFs were observed, i. e., inhibition (39/44, 88.6%), facilitation (2/44, 4.6%), and no effectiveness (3/44, 6.8%). Statistical analysis indicated that only inhibitory effect of weak noise was significant (P< 0.001, n = 39). The inhibitory effect of weak noise was greater at lower stimulus intensity of tone burst but decreased significantly with increased stimulus intensity (P< 0.0001, n = 39). In addition, the DR and slope of RIFs became narrower and steeper with weak noise presentation, respectively (P< 0.01, n = 31). The results from the present study suggest that weak noise exerts a dynamic modulatory action on acoustical intensity sensitivity of IC neurons, which possibly leads to a better understanding of neural mechanisms underlying the extraction of sound signals from natural auditory scenes.