Temporal Response Properties Of Auditory Neurons To Sound Amplitude In Mouse Central Nervous System

Encoding of sensory information by the timing of neural discharges in central nervous system has received more and more agreements. The timing of spikes, especially the first spike latency (FSL) has received increasing attention as a more precise and reliable information carrier. The evidences have been uncovered in several neural systems including visual, somatosensory, olfactory, hippocampus and auditory systems. The FSL coming up almost systematically as a highly reliable coding dimension might turn out to be the general unit of sensory representation.Most neurons in the auditory pathway discharge spikes locking to the onset of an acoustic stimulus and maintained a sober state of alertness to the outside world even in anesthesia, coma, sleep and other state. The auditory system has its own advantages as the research model for the temporal coding mechanism. The basic parameters of acoustic stimulus include the sound stimulus frequency, intensity and sound duration. In the previous studies, pure tone bursts with a a single frequency and amplitude were used as the acoustic stimuli.Generally, this temporal envelope is experimentally controlled using four static parameters (rise time, RT; steady-state pressure, SP; sound duration, D; fall time) and two dynamic parameters (rise function, RF and fall function; routinely, liner or cosine-squared used). The peak pressure(PP) of onset of the stimulus temporal envelope was shaped by the RT and RF and the PP of offset was shaped by the FT and FF.Spike counts (SC), spike rate(SR) and FSL were used to evaluate the responses of neurons to the amplitudes and frequencies of acoustic stimuli. It was clear that FSL with its precision and stability appeared to be a better parameter than SC in the evaluation of the response of a mouse inferior colliculus neuron to the frequency and amplitude of acoustic stimulus when its duration was fixed. However, the representation of sound duration by FSL and the influence of sound duration on the neural responses to the other two parameters were unknown.In this study, we employed extracellular recording technique on central nuclei of inferior colliculus in 4 to 6 weeks BALB/c mice to explore the affection of sound duration on FSL, an SP-D scan was performed with various stimulus parameters. SC and FSLs were both used to evaluate the neuronal responses to the binaural acoustic stimuli which were varied with duration and amplitude at the characteristic frequency.The FSL-versus-D and-versus-SP, SC-versus-D and -versus-SP were plotted to analyze the effect of sound duration on the representation of acoustic amplitude as well as the relationship between FSL and duration.A total of 34 neurons were recorded. The results showed that the SC-amplitude functions across a population of auditory neurons were classified as the monotonic (32%), saturated (27%) and non-monotonic (41%). For each type, the iso-duration SC-amplitude curves showed similar shapes with large variations. In contrast to SC, FSL decreased monotonically with the amplitude increasing. The iso-duration FSL-amplitude curves of a recorded neuron overlapped. It revealed that sound duration had no effect on the representation of acoustic amplitude; FSL could precisely reflect the response to acoustic amplitude and its variation. It was also observed that the SC changing with the sound duration and SC-duration functions were unstable with the changes of amplitude, however, the FSLs of all neurons did not change with the duration for a given sound intensity.Conclusion:1) Sound duration has no effect on the representation of acoustic amplitude;2)It is further improved of the precision and stability of FSL to be a better parameter than spike counts in evaluation of the response of a neuron to acoustic amplitude and its variation:FSL decreases as the amplitude increases. Thus, we believe that sound duration and spike counts have no effect on the FSL.The temporal envelope was experimentally controlled using static parameters and dynamic parameters. So what parameters of sound stimulus onset contribute to the FSL of neuronal responses? It remains somehow ambiguous. A number of previous studies have shown that FSLs in most auditory neurons were determined by the onset of the stimulus temporal envelope and shaped by the primary stimulus parameters, which determine the envelope onset status, such as the SP, RT and RF. The researchers analyzed the experimental data by models. In the previous study, we designed a series of exponential RFs (t/RT)n varied with the exponent (n). We compared the responses of inferior colliculus of mice to the acoustic stimuli with different RT, RF, and SP. FSL is not found in response to changes in the speed and acceleration of sound intensity, etc., but is equal to the temporal integral of the instantaneous sound peak pressure,that is,(?) PP dt.T is the neuron firing threshold and K=2. In order to further prove K=2 is suitable for auditory central nervous system by this encoding mechanism of neurons, in this present study, we employed extracellular recording technique on auditory cortex to explore the response of neurons. In the SP-RT-RF scan experiment, we changed the temporal envelope shaping pure tone busts by altering the values of SP, RT and RF, programmed by RPvds software, in which when n=1,2, or 3 it is equivalent to SP×(t/RT)(linear,L), SP×(t/RT)2 (quadratic, Q) and SP×(t/RT)3(cubic, C) individually. The FSLs of a given neuron responding to an alteration in the experimentally-controlled sound parameters (SP, RT and RF or n) were sorted into three groups (FSLn) obtained using the n=1,2 and 3 RF tones. And the FSL-versus-SP and-versus-RT and-versus-SP/RTn were plotted. Therefore, FSL-SP/RTn functions were fit to the equation. MFSL and T are two free parameters. R2, MFSL and lgT value obtained by curve-fitting. Histograms of R2 value at RFs of all neurons were plotted. In order to observe the effect of RFs on MFSL and T, we analyzed and compared MFSL values and lgT values of the neurons at three RFs. Repeated-measures were applied to test the difference of the results. The correlation between lgT and MT across the population of neurons were plotted in OriginPro7.5.A total of 25 well-isolated neurons from auditory cortex were obtained. FSL-versus-SP,-versus-RT, and-versus-SP/RTn were plotted, which showed that the FSL decreased as steady-state pressure increased. However, the FSL increased with an increasing rise time. When K=2, the FSL-invariant-coefficient functions were fitted well. Histograms of the distribution of R2 values showed the fitting rates were above 0.91 for 20 neurons (n=1) and 24 neurons (n=2) and above 0.925 for 21 neurons(n=3). The values of MFSL and T of 22 neurons showed no clear relationship with the three RFs. The differences among MFSL or lgT of 22 neurons were not statistically significant at three RFs(Repeated measures ANOVA, MFSL values comparisons:F=2.375, P=0.138; T values comparisons:F=2.464, P=0.131). The values of MFSL and T for different neurons are not the same, but for same neurons they are fixed, which are not affected by the RFs. And T also showed a good linear correlation with the neuronal MT.Conclusion:1) When K= 2, the sound intensity integral model is appropriate to the auditory cortex.2) T and MFSL are constant for a neuron, but different form neurons.T is a mirror of MT.3) The results obtained from cortical neurons are consistent with those in inferior colliculus in our previous research, which provides evidence for this kind of coding mechanism may be general for all neurons in auditory central nervous system, and a new animal model.