Learning is acquiring information and memory is storing it. A common belief is that sensory systems are only "sensory analyzers" that provide instantaneous knowledge of the world, but are not sites of information storage. However, during the last decade of the twentieth century, neurophysiological recordings in the auditory cortex revealed that learning and memory involve specific and enduring changes in the way that the brain analyzes and stores sound. This article covers the topic of neural plasticity in the auditory cortex (ACx) during learning and memory in adult animals and humans. As used here, the term "neural plasticity" refers to systematic, long-term (minutes to lifetimes) changes in the responses of neurons to the same physical stimulus (e.g, a tone) arising out of an experience.
In 1955 Galambos and his colleagues performed a seminal experiment in which cats were classically conditioned by pairing an auditory conditioned stimulus (CS) with a puff of air (unconditioned stimulus, or US) to the face. Evoked potentials elicited by the CS in the ACx became larger during conditioning. This was soon followed by a two-tone discrimination study showing that the increased magnitude of response was caused by the specific association of a CS and US versus a second, unpaired tone. Many more studies in various animals and training situations also demonstrated that the responses of the auditory cortex to sounds were affected by the learned psychological importance of acoustic stimuli.
But these results had little effect on the neurobiology of learning/memory or auditory neurophysiology. Auditory physiologists took scant notice of learning studies that used only one or two tones because they shed little light on how the ACx processed the many acoustic frequencies to which it is sensitive. Learning/memory workers persisted in the shared assumption that the auditory cortex is a "sound analyzer" but not a "learning machine." But this assumption is not corroborated by the brain.
Receptive Field Plasticity in Learning and Memory
A solution to the impasse involved combining methods from auditory neurophysiology with those from learning and memory. Researchers determined the effects of classical conditioning on the receptive fields of neurons in the ACx. The receptive field (RF) of a cell is the part of the stimulus environment to which it is sensitive. Thus, the RF for frequency is described by a "tuning curve" that plots the number of cellular discharges as a function of acoustic frequency. The "best frequency" (BF) of a cell is the frequency that elicits the most discharges. Demonstrating that learning involves a systematic change in the processing of acoustic-frequency information involves showing an alteration in receptive fields when a subject learns that a particular frequency is behaviorally important.
This approach was first applied to the ACx in 1990 at the University of California at Irvine. RFs were obtained before and immediately after a single conditioning session in which a tone was paired with a mild shock. Behavioral conditioned responses developed. More importantly, the RFs developed systematic plasticity. Responses to the CS frequency increased, whereas responses to the best frequency and many other frequencies decreased. These opposing changes were often sufficient to shift frequency tuning toward or to the frequency of the CS so that it became the new BF (see Figure 1). Tuning shifts were caused only by association between a tonal CS and the shock US. Also, this plasticity was not caused by arousal to the CS frequency because RF shifts are still observable after training when subjects are placed under general anesthesia and thus not aroused by stimuli.
RF plasticity possesses all of the major characteristics of memory. In addition to being associative, it is highly specific to the CS frequency, develops very rapidly (within five trials), exhibits consolidation (i.e., becomes stronger without additional training), and is retained indefinitely (tested to eight weeks). Moreover, RF plasticity exhibits generality across different types of training (e.g., one-and two-tone discrimination training, in both in classical and instrumental conditioning), and motivation (appetitive as well as aversive). RF plasticity also develops in humans and produces an expanded representation of the CS frequency in the ACx.
RF plasticity may provide a memory code for acquired stimulus importance: the greater the behavioral significance, the greater the number of cells that become tuned to that stimulus. Such a memory code could explain aspects of selective attention—for example, why one is more likely to hear one's own name in a noisy room than a random name. While background noise may control many neurons, the attunement of still more more networks of neurons to our names increases the probability that the name will engage some of these cells. This process could explain why the loss of memory in aging and brain degeneration take less of a toll on one's most important memories. If the memory is represented by more cells, then important information has a "safety factor" in numbers, which blunt the impact of cell loss.
Mechanisms and Functions
Subcortical neuromodulatory transmitter systems probably play a role in ACx plasticity in learning. The preponderance of evidence implicates the nucleus basalis cholinergic system (NB/ACh) in enabling the auditory cortex to store specific information. For example, RF plasticity induced in a variety of tasks, motivations, and species also can be induced by substituting NB stimulation for appetitive or aversive reinforcement. That this plasticity can be blocked by the administration of atropine directly to the ACx shows that the induced RF plasticity requires the engagement of cholinergic muscarinic receptors in the ACx. Moreover, pairing a tone with stimulation of the NB/ACh is sufficient to induce actual specific behavioral memory in the rat (see Figure 2). Although much more research is necessary, researchers have established the outlines of a reductionistic account of learning as the storage of specific information in the auditory cortex.
Bakin, J. S., and Weinberger, N. M. (1990). Classical conditioning induces CS-specific receptive field plasticity in the auditory cortex of the guinea pig. Brain Research 536, 271-286.
Galambos, R., Sheatz, G., and Vernier, V. G. (1955). Electrophysiological correlates of a conditioned response in cats. Science 123, 376-377.
Gonzalez-Lima, F., and Scheich, H. (1986). Neural substrates for tone-conditioned bradycardia demonstration with 2-deoxyglucose: II. Auditory cortex plasticity. Behavioral Brain Research 20, 281-293.
Gluck, H., and Rowland, V. (1957). Defensive conditioning of electrographic arousal with delayed and differentiated auditory stimuli. Electroencephography and Clinical Neurophysiology 11, 485-491.
McLin, D. E. III, Miasnikov, A. A., and Weinberger, N. M. (2002). Induction of behavioral associative memory by stimulation of the nucleus basalis. Proceedings of the National Academy of Sciences of the United States of America.
Morris, J. S., Friston, K. J., and Dolan, R. J. (1998). Experience-dependent modulation of tonotopic neural responses in human auditory cortex. Proceedings of the Royal Society of London 265, 649-657.
Recanzone, G. H., Schreiner, C. E., and Merzenich, M. (1993). Plasticity in the frequency representation of primary auditory cortex following discrimination training in adult owl monkeys. Journal of Neuroscience 13, 87-103.
Weinberger, N. M. (2002). Experience-dependent response plasticity in the auditory cortex: Issues, characteristics, mechanisms and functions. In T. Parks, ed., Handbook of Auditory Research. New York: Springer-Verlag.
Weinberger, N. M., and Diamond, D. M. (1987). Physiological plasticity of single neurons in auditory cortex: Rapid induction by learning. Progress in Neurobiology 29, 1-55.