PubMed2000
Recordings of activity were made from 647 single units of the A(I) cortex of awake cats to evaluate behavioral state-dependent changes in the population response to a 70-dB click. Averages of PST histograms of unit activity were used to assess the changes in response. This report focuses on slow components of the responses disclosed by averages employing bin widths of 16 ms. Responses were compared before and after a Pavlovian blink CR was produced by forward pairing of click conditioned stimuli (CSs) with USs. A backward-paired 70-dB hiss was presented as a discriminative stimulus. Studies were also done after backward pairing of the click CSs (backward conditioning) that produced weak sensitization instead of a conditioned response. There were four main findings. First, components of activity elicited 32-160 ms after presenting the hiss decreased significantly after conditioning and after backward conditioning. The decreases after conditioning represented the most pronounced changes in activity evoked by either clicks or hisses in this behavioral state. Second, baseline firing decreased after both conditioning and backward conditioning. The direction of baseline change was opposite that found in adjacent cortical regions and in A(I) cortex after operant conditioning employing an acoustic cue. Third, prior to conditioning, unit activity in response to the hiss declined before the sound of the hiss reached its peak or terminated. This decrease was thought to represent a habituatory adaptation of response to a prolonged acoustic stimulus. This type of habituation to a lengthy stimulus has been recognized, behaviorally, but has not been observed previously in the activity of units of the auditory receptive cortex. Fourth, the percentage of click responsive units did not change significantly after the click was used as a CS for conditioning, and despite the accompanying changes in baseline activity, the absolute levels of activity summed in the first 16 ms after click delivery remained stable across behavioral states in which the motor response to the click was altered profoundly. The onset of the conditioned motor response began 20 ms after the click, and was shown earlier to depend on rapid, potentiated transmission through the cochlear nucleus and motor cortex for its generation. Thus the stability of the response to the click in the primary auditory receptive cortex was unexpected. This led us to make further analyses of the data with 2- and 4-ms bin widths (see companion report) that eventually disclosed a potentiated response to the click. The findings show stability and change in the response to the click as a CS, depending on the band pass (bin width) used for analysis of spike activity. In the representation disclosed by low pass filtering in this study, the response was stable. This representation provided information suitable for identifying commonalties of the click signals across varying behavioral states. The representations of the click and hiss contained in the slow components of the population response in the A(I) cortex were uncorrelated with the selective potentiation of activity in motor cortex and behavioral performance in response to click as a CS after conditioning. Although changes in the activity evoked by hisses occurred after conditioning, the changes also occurred after backward conditioning when only small, sensitized behavioral responses to clicks and hisses were observed. Basic theoretical considerations about information transmission in complex neural networks plus clinical observations comparing derangements of linguistic and non-linguistic cortical functions in humans suggest that multiple representations of conditioned stimulus inputs may exist in local populations of cortical neurons. Together, our studies provide evidence for two different, concurrent representations of information about a click CS encoded in the spike activity of the A(I) cortex.