Norman M. Weinberger

Specific long-term memory traces in primary auditory cortex

Learning and memory involve the storage of specific sensory experiences. However, until recently the idea that the primary sensory cortices could store specific memory traces had received little attention. Converging evidence obtained using techniques from sensory physiology and the neurobiology of learning and memory supports the idea that the primary auditory cortex acquires and retains specific memory traces about the behavioural significance of selected sounds. The cholinergic system of the nucleus basalis, when properly engaged, is sufficient to induce both specific memory traces and specific behavioural memory. A contemporary view of the primary auditory cortex should incorporate its mnemonic and other cognitive functions.

Classical conditioning induces CS-specific receptive field plasticity in the auditory cortex of the guinea pig

To determine if classical conditioning produces general or specific modification of responses to acoustic conditioned stimuli (CS), frequency receptive fields (RF) of neurons in guinea pig auditory cortex were determined before and up to 24 h after fear conditioning. Highly specific RF plasticity characterized by maximal increased responses to the CS frequency and decreased responses to the pretraining best frequency (BF) and other frequencies was observed in 70% of conditioning cases. These opposing changes were often sufficient to produce a shift in tuning such that the frequency of the CS became the new BF. CS frequency specific plasticity was maintained as long as 24 h. Sensitization training produced general increased responses across the RF without CS specificity. The findings indicate that associative processes produce systematic modification of the auditory systems processing of frequency information and exemplify the advantages of combining receptive field analysis with behavioral training in the study of the neural bases of learning and memory.

Physiological plasticity in auditory cortex: Rapid induction by learning

Abbreviations 2

Is the Amygdala a Locus of “Conditioned Fear”? Some Questions and Caveats

freezing and FPS (Walker and Davis, 1997; see also Since the publication of Weiskrantz’ seminal paper, widely Davis, 1997, for a review of much of the relevant literainterpreted as suggesting that amygdala lesions impair ture). Therefore, critical additional evidence must be the formation of stimulus–reinforcement associations provided to exclude the very real possibility that the (Weiskrantz, 1956), extensive research in many laboraimpairing effects of L/BL lesions are due simply to an tories has attempted to determine the role(s) of the impaired ability of rats to express freezing or FPS, amygdala in learning and memory. In recent years, many whether it is conditioned or unconditioned. Clearly, until studies have investigated amygdala participation in such evidence is provided, it is premature to conclude Pavlovian “fear conditioning” in rats. The findings of on the basis of results of lesion studies that the L/BL is such studies have suggested that the amygdala may be necessary for the acquisition and expression of condian essential link in a subcortical circuit mediating the tioned fear. formation and permanent (or “indelible”) storage of “condiThe second issue concerns interpretation of the eftioned fear” (reviewed by LeDoux, 1995; Maren and fects of intraamygdala infusions of NMDA antagonists Fanselow, 1996; Davis, 1997). Succinctly put, interpretaon Pavlovian “fear conditioning.” Several studies have tions of the findings have suggested that “...attaching reported that Pavlovian conditioned freezing or FPS is ‘fear’ to the previously neutral stimulus and rememattenuated by NMDA antagonists infused into the amygbering it is what the amygdala does” (Stevens, 1998). dala (directed at the L/BL) prior to conditioning (see Evidence from many studies is consistent with the hyLeDoux, 1995; Maren and Fanselow, 1996; Davis, 1997). pothesis that the amygdala—in particular, the lateral/ Such findings are consistent with the hypothesis that basolateral (L/BL) nuclei—encodes and permanently induction of NMDA-dependent long-term potentiation stores learned fear. However, there is also evidence that (LTP) within the L/BL may be the neural substrate of challenges this hypothesis. Evaluation of the hypothesis “conditioned fear” (Rogan et al., 1997). However, in and the supporting evidence requires consideration and entertaining this hypothesis, it is essential to provide clarification of several critical issues. evidence that excludes non-mnemonic effects of NMDA The first issue concerns interpretation of the effects antagonists that might disrupt acquisition, such as efof L/BL lesions on the expression of “conditioned fear.” fects on attentional or motivational processes. FurtherIt is well established that L/BL lesions impair the expresmore, and more importantly, an unambiguous demonsion of conditioned fear, indexed either by “freezing” in stration that intra-L/BL infusions of NMDA antagonists the presence of a cue previously paired with footshock impair learning would not address the critical tenet of or by startle to a tone in the presence of a cue previously the hypothesis that the L/BL is the permanent storage paired with footshock (fear-potentiated startle or FPS) site for “conditioned fear.” Such findings would be com(LeDoux, 1995; Maren and Fanselow, 1996; Davis, 1997). pletely consistent with the possibility that activation of Indeed, excitotoxic lesions of the L/BL impair freezing NMDA receptors in the L/BL influences short-term memand FPS even when induced 1 month after the conditionory processes within the L/BL and/or alters L/BL activity ing (Lee et al., 1996; Maren et al., 1996). Such impairregulating long-term memory storage in other brain rements are generally interpreted as evidence that L/BL gions (McGaugh et al., 1984; Cahill and McGaugh, 1998). lesions block conditioned fear. However, when using A final issue concerns interpretation of electrophysiolesions to study the role of a brain region in memory, it logical changes in the L/BL associated with “fear condiis absolutely essential to distinguish the lesion’s effects tioning.” Recent findings suggest, for example, that LTP on memory from other influences on performance (the develops in the L/BL during Pavlovian “fear conditionwell-known “learning/performance” distinction). The coning” (Rogan et al., 1997). However, it is well established clusion that a brain lesion impairs memory requires evithat, within regions of the medial geniculate nucleus dence that the lesion does not simply disrupt the aniknown to project to these same amygdala regions, Pavmal’s ability to make the specific response(s) that are lovian “fear conditioning” induces specific, associative used as evidence of memory. Thus, Richard Thompson changes in neuronal responses to the CS used to induce and colleagues (Thompson et al., 1998), for example, conditioning (for discussion, see Ryugo and Weinbegan their investigation of the role of the deep cerebelberger, 1978; Weinberger, 1998). Additionally, converlar nuclei in Pavlovian eyelid conditioning by demongence of auditory and noxious sensory stimulation is

Rapid development of learning-induced receptive field plasticity in the auditory cortex.

Classical conditioning induces frequency-specific receptive field (RF) plasticity in the auditory cortex after relatively brief training (30 trials), characterized by increased response to the frequency of the conditioned stimulus (CS) and decreased responses to other frequencies, including the pretraining best frequency (BF). This experiment determined the development of this CS-specific RF plasticity. Guinea pigs underwent classical conditioning to a tonal frequency, and receptive fields of neurons in the auditory cortex were determined before and after 5, 15, and 30 CS-US (unconditioned stimulus) pairings, as well as 1 hr posttraining. Highly selective RF changes were observed as early as the first 5 training trials. They culminated after 15 trials, then stabilized after 30 trials and 1 hr posttraining. The rapid development of RF plasticity satisfies a criterion for its involvement in the neural bases of a specific associative memory.

Receptive field plasticity in the auditory cortex during frequency discrimination training: selective retuning independent of task difficulty.

Classical conditioning is known to induce frequency-specific receptive field (RF) plasticity in the auditory cortex (ACx). This study determined the effects of discrimination training on RFs at two levels of task difficulty. Single unit and cluster discharges were recorded in the ACx of adult guinea pigs trained in a tone-shock frequency discrimination paradigm (30 intermixed trials each of positive conditioned stimulus [CS+]-shock and negative CS [CS-] alone) with behavioral performance indexed by the cardiac deceleration conditioned response (CR). After training in an easy task in which subjects developed discriminative CRs, they were trained in a difficult task (reduced frequency distance between CS+ and CS-) in which they failed to discriminate. However, frequency-specific RF plasticity developed at both levels of task difficulty. Responses to the frequency of the CS+ were increased, whereas responses to other frequencies, including the CS- and the prepotent best frequency (BF) were reduced. In many cases, tuning was shifted such that the frequency of the CS+ became the new BF. The effects were present or stronger after a 1-hr retention interval. The role of RF plasticity in the ACx is discussed for behavioral performance and information storage.

Associative retuning in the thalamic source of input to the amygdala and auditory cortex: receptive field plasticity in the medial division of the medial geniculate body.

The medial division of the medial geniculate body (MGm) projects to the lateral amygdala and the upper layer of auditory cortex and develops physiological plasticity rapidly during classical conditioning. The effects of learning on frequency receptive fields (RFs) in the MGm of the guinea pig have been determined. Classical conditioning (tone-footshock), as indexed by rapid development of conditioned bradycardia, produced conditioned stimulus (CS)-frequency specific RF plasticity: increased response at the CS frequency with decreased responses at other frequencies, both immediately and after a 1-hr retention period. Sensitization training produced only general changes in RFs. These findings are considered with reference to both the elicitation of amygdala-mediated, fear-conditioned responses and the mechanism of retrieval of information stored in the auditory cortex during acquisition.

Classical conditioning rapidly induces specific changes in frequency receptive fields of single neurons in secondary and ventral ectosylvian auditory cortical fields

To determine if learning-induced changes in the response of auditory cortical neurons to a conditioned stimulus (CS) reflect general changes in cellular excitability or alterations in signal processing that are specific to that stimulus, we determined frequency receptive fields (FRFs) of single neurons in secondary and ventral ectosylvian auditory fields of the cat during classical conditioning. Associative changes in FRFs of most cells were specific to the frequency of the CS, established rapidly and reversed by extinction. Thus, learning causes specific changes in cortical processing of sounds whose significance is acquired.

Physiological plasticity of single neurons in auditory cortex of the cat during acquisition of the pupillary conditioned response: I. Primary field (AI).

The discharges of 22 single neurons were recorded in the secondary auditory cortical field (AII) during acquisition of the pupillary dilation conditioned defensive response in chronically prepared cats. All 22 neurons developed discharge plasticity in background activity, and 21/22 cells developed plasticity in their responses to the acoustic conditioned stimulus (CS). Nonassociative factors were ruled out by the use of a sensitization phase (CS and US [unconditioned stimulus] unpaired) preceding the conditioning phase and by ensuring stimulus constancy at the periphery by neuromuscular paralysis. Changes in background neuronal activity were related to measures of behavioral learning or to changes in the level of arousal. Specifically, decreases in background activity (17/22 cells) developed at the time that subjects began to display conditioned responses. Increases in background activity (5/22) developed in animals that became more tonically aroused during conditioning. However, both increases (11/22) and decreases (10/22) in evoked activity developed independently of the rate of pupillary learning, tonic arousal level, or changes in background activity. These findings indicate that changes in background activity are closely related to behavioral processes of learning and arousal whereas stimulus-evoked discharge plasticity develops solely as a consequence of stimulus pairing. A comparative analysis of the effects of conditioning on secondary and primary (AI) auditory cortex indicates that both regions develop neuronal discharge plasticity early in the conditioning phase and that increases in background activity in primary auditory cortex are also associated with elevated levels of tonic arousal. In addition, the overall incidence of single neurons developing learning-related discharge plasticity is significantly greater in AII than in AI. The relevance of these findings is discussed in terms of parallel processing in sensory systems and multiple sensory cortical fields.

The neurobiology of learning and memory: some reminders to remember

We have learned much about the neurobiology of learning and memory in the past 100 years. We have also learned much about how we should, and should not, investigate these complex processes. However, with the rapid recent growth in the field and the influx of investigators not familiar with this past, these crucial lessons too often fail to guide the research of today. Here we highlight some major lessons gleaned from this wealth of experience. These include the need to carefully attend to the learning/performance distinction, to rely equally on synthetic as well as reductionistic thinking, and to avoid the seduction of simplicity. Examples in which the lessons of history are, and are not, educating current research are also given.