James L. Ringo - US grants

Investigation of visual memory in the nonhuman primate has enjoyed remarkable recent success, among other things, in demonstrating the closeness between human and nonhuman primate mnemonic processing and amnesias. This success has come largely from using the technique of destruction of relevant portions of the neuronal substrate, e.g., hippocampal formation. However, possibly because of limitations of lesioning technique, information on the time course of initial storage and retrieval occurring within the putative mnemonic circuitry is absent in this account. Experiments performed in this laboratory, during the previous support period, have now revealed a unique opportunity to address such questions with unexpected precision. It was found that a single electrical burst of one to four pulses lasting less than 50 milliseconds (ms), delivered to the hippocampal formation in macaques 50-200 ms after a 50 ms view of visual image, eliminates their ability to recognize that image subsequently. Such pulses before or after this period are essentially without effect. Similarly, delivery of electrical stimulation while the animal is endeavoring to retrieve a previously viewed image severely disrupts its accuracy. Such pulses are otherwise without effect upon performance. The proposed experiments then, will use brief perturbing pulses to measure the temporal course of mnemonic processing with great precision, and do so for different components of the hippocampal formation. to this end, advantage will be taken of the fact that performance by a single hemisphere in a split-brain macaque is essentially equivalent to that with both hemispheres; thus it will be possible to achieve the perturbation from a single locus at a time, greatly enhancing the ability to specify critical sites. The contribution of this project is aimed directly at the neuronal circuitry supporting memory. Its clinical relevance is not immediate but could nonetheless be profound since it would be rather surprising if a full understanding of the mnemonic circuitry did not have a major impact on the important - and with the aging population - growing health problems involving memory.

Investigation of visual memory in the nonhuman primate has enjoyed remarkable recent success, among other things, in demonstrating the closeness between human and nonhuman primate mnemonic processing and amnesias. This success has come largely from using the technique of destruction of relevant portions of the neuronal substrate, e.g., hippocampal formation. However, possibly because of limitations of lesioning technique, information on the time course of initial storage and retrieval occurring within the putative mnemonic circuitry is absent in this account. Experiments performed in this laboratory, during the previous support period, have now revealed a unique opportunity to address such questions with unexpected precision. It was found that a single electrical burst of one to four pulses lasting less than 50 milliseconds (ms), delivered to the hippocampal formation in macaques 50-200 ms after a 50 ms view of visual image, eliminates their ability to recognize that image subsequently. Such pulses before or after this period are essentially without effect. Similarly, delivery of electrical stimulation while the animal is endeavoring to retrieve a previously viewed image severely disrupts its accuracy. Such pulses are otherwise without effect upon performance. The proposed experiments then, will use brief perturbing pulses to measure the temporal course of mnemonic processing with great precision, and do so for different components of the hippocampal formation. to this end, advantage will be taken of the fact that performance by a single hemisphere in a split-brain macaque is essentially equivalent to that with both hemispheres; thus it will be possible to achieve the perturbation from a single locus at a time, greatly enhancing the ability to specify critical sites. The contribution of this project is aimed directly at the neuronal circuitry supporting memory. Its clinical relevance is not immediate but could nonetheless be profound since it would be rather surprising if a full understanding of the mnemonic circuitry did not have a major impact on the important - and with the aging population - growing health problems involving memory.

Work in the previous support period led to a finding that electrical stimulation of the medial temporal lobe of monkeys during the delay period of a delayed matching-to-sample task was highly disruptive of performance if only a few images were repeatedly reused. The identical stimulation was only mildly disruptive if fresh images were used for each trial. In memory tasks with human subject, the use of repeated images leads to a strategy of active rehearsal during the delay period. In monkeys, the use of repeated images is a key to inducing the so- called delay activity. This is an elevated discharge rate in neurons usually lasting throughout the delay period, with different neurons selectively active for different stimuli. Thus, the activity across the population of neurons encodes a memory of that which is to-be- remembered. The delay activity appears like a neuronal version of continuous rehearsal, and is an excellent candidate for the neural embodiment of memory during the delay under these conditions. The fascination with this delay activity is that it may be our best model for the everyday human working memory which is the substrate for thought. The evidence that delay activity is the holder of the memory is, however, indirect. The purpose of the current proposal is to test that hypothesis directly. This will be done by using electrical stimulation which is known to disrupt that delay activity (see Preliminary Studies). If delay activity of neurons is the carrier of the memory then disruption of the delay activity should disrupt the behavioral memory. The contribution of this project is aimed directly at a prominent candidate neuronal mechanism of memory. Its clinical relevance is not immediate but could nonetheless be profound since a full understanding of the mnemonic circuitry should have a major impact on the important - and with the aging population - growing health problems involving memory.

DESCRIPTION: Recognition memory is a simple form of memory allowing well controlled experimental investigation. In the prototypical experiment, two separate presentations are made of a stimulus. In the first presentation the stimulus is novel and in the second it is familiar. In the last 25 years many laboratories have performed this type of experiment, seeking evidence of mechanisms underlying recognition by recording the spike trains from neurons in high level visual and association cortex of primates. One prominent candidate mechanism for recognition memory has been readily recorded under such circumstances, a decrement in response amplitude specific to previously presented images, called here a stimulus-specific adaptation (SSA). This SSA is robust; it has been recorded in many laboratories and is clearly evident in both single units and population averages from many areas (inferotemporal, perirhinal, entorhinal, V4, prefrontal cortex). Except under special behavioral circumstances, this is the only candidate mechanism which has been found for recognition memory in primates, and it is widely hypothesized to be the mediator of recognition memory. It was thus a surprise when the applicant was recently able to eliminate SSA from the neural responses of units recorded in inferotemporal cortex, without a concomitant elimination of behavioral recognition. The adaptation was eliminated by a using a unique, partial split-brain preparation in which two completely separate routes were used to bring information to the recorded units. Despite this result, SSA is still the leading, and currently really the only, candidate mechanism for recognition memory. The best way to reconcile this result with the hypothesis that SSA is the mechanism underlying recognition memory is to suggest that while the adaptation could be eliminated in inferotemporal cortex without eliminating behavioral recognition, this could not occur one or two stages further into higher level association cortex. To test this hypothesis, a series of experiments is proposed to determine whether SSA and behavioral recognition can be dissociated in entorhinal, perirhinal, and/or ventral prefrontal cortex.