The Cerebral Cortex for the Adjustment and Improvement of Auditory Signal Processing

Wednesday, March 10, 1999 - 9:00am - 10:00am
Keller 3-180
Nobuo Suga (Washington University)
In the mustached bat (Pteronotus parnellii), cortical auditory neurons mediate, via corticofugal projection, a highly focused positive feedback to subcortical neurons matched in tuning to a particular acoustic parameter in the frequency or time domain, and a widespread lateral inhibition to unmatched subcortical neurons. This cortical feedback results in the adjustment and improvement of subcortical signal processing (i.e., the adjustment and improvement of the cortical neurons' own input). This function, named egocentric selection, enhances the neural representation of frequently occurring signals in the central auditory system (Yan & Suga 1996; Zhang, Suga & Yan 1997). In the big brown bat (Eptesicus fuscus), egocentric selection shifts the best frequencies of collicular neurons not only toward the best frequency of electrically stimulated cortical neurons but also towards the frequency of a repetitively delivered acoustic stimulus (tone burst), resulting in local reorganization of the frequency map in the subcortical nuclei (Yan & Suga 1998), and also the auditory cortex (Chowdhury & Suga 1998). It also evokes such reorganization according to auditory experience based on associative learning (Gao and Suga, 1998).

The effect of egocentric selection is apparently different between the two species of bats studied and, perhaps, between different portions of a frequency map of the central auditory system of the mustached bat, reflecting the shape and sharpness of frequency-tuning curves. Therefore, I may propose the hypothesis that egocentric selection is a basic neural mechanism shared by many mammalian species, but its effects on subcortical neurons, and accordingly, cortical neurons, can be somewhat different from species to species, depending on species-specific needs for auditory signal processing.

Gao and Suga (1998) proposed the following model, which is somewhat different from Weinberger's model (1998) based on LeDoux and Muller's model (1997). A train of acoustic stimuli (AS) and an electric leg-stimulation (ES) excite the auditory and somatosensory cortices, respectively. These sensory cortices send signals to the amygdala through the association cortex. AS evokes changes in the auditory cortex, which are highly specific to acoustic stimuli and are based on egocentric selection working together with the corticofugal (descending) system. (The corticofugal system forms a feedback loop with the ascending system.) When associative learning takes place in the amygdala, that is, when an animal is conditioned by AS followed by ES, the cholinergic basal forebrain is excited by AS + ES through the amygdala and increases the acetylcholine level in the cortex. As a result, the AS-related changes in the auditory cortex are augmented in magnitude and duration. In other words, the processing of acoustic stimuli is adjusted and improved according to the behavioral relevance of the stimuli.

I will try to simplify my story to be understandable to non-auditory physiologists. (Work supported by NIDCD DC 00175).