Does the consolidation period have psychological utility

The memory/LTP hypothesis predicts that naturally occurring patterns of brain activity, as well as patterns related to learning in particular, induce the potentiation effect. The discovery that this is in fact the case (Larson and Lynch, 1986; Larson, Wong, and Lynch, 1986) provided one of the more convincing links between memory and potentiation. The pertinent work has been reviewed elsewhere and will only be noted here as necessary background to a seldom-asked question about consolidation posed by studies on LTP reversal.

A 4-7 Hz rhythm (theta) occurs throughout the cortical telencephalon and is particularly prominent in the hippocampus of small mammals during active behavior. Individual cells commonly emit short bursts of high-frequency activity (3-4 pulses in 30-50 ms) with the bursts being synchronized with the rhythm (Otto et al., 1991). These conditions (bursts separated by 150-250 ms) proved to be ideal for inducing LTP when used as a stimulation pattern. Studies on the reasons for this result showed that theta takes advantage of an IPSP refractory process that reaches its maximum at about 200 ms (Larson, Wong, and Lynch, 1986). Theta thus momentarily releases target neurons from feedforward inhibition (Larson and Lynch, 1986), allowing a second theta burst to generate depolarization sufficient to activate NMDA receptors (Larson and Lynch, 1988) and produce the elevations of internal calcium needed to induce potentia-tion. Other workers found that the refractory period itself is due to inhibition of GABAergic terminals by GABAb autoreceptors (Mott and Lewis, 1991). In all, the investigations of the LTP-theta relationship provided a first success in coupling brain rhythms to synaptic chemistry. Moreover, the connection of patterns of brain activity with increments in synaptic strength provided mechanistic explanations for fundamental features of memory including the prominence of temporal contiguity in the forming of associations.

The fact that LTP induction has a deep relationship with a particular rhythm inevitably raised the question of whether this is also true for LTP reversal. Confirmation came in a series of experiments using hippocampal slices (Larson, Xiao, and Lynch, 1993). While the existence of particularly appropriate patterns for reversal was not unexpected, it was indeed surprising that the pertinent rhythm proved to be theta. This discovery strongly implies that erasure requires the prolonged EPSPs that result from IPSP suppression, presumably for altogether different reasons than those involved with induction. Possible mechanisms suggested by theta and other aspects of reversal will be taken up briefly in a later section. But independent of neurobiological issues, the connection of reversal to a prominent rhythm points to the possibility that it is a "real world" event, that is, one that occurs routinely in behaving animals. This impression is strengthened by a more detailed examination of the experiments in which the connection was found. Two inputs (groups of Schaffer-commissural fibers) that converged on a target population of dendrites were used (figure 10.7). Pairs of theta bursts were applied simultaneously to each input; 20 seconds later, single pulses at 5 per second (theta) were delivered to one of the two inputs for 30 seconds. This procedure was repeated every 2 minutes for a total of five pairs of theta bursts. Absent theta pulse stimulation, both inputs increment their syn-aptic strength after each pair of theta bursts until a ceiling is reached after 5-10 pairs; with the pulses, one input grows stronger and the other does not change (see Larson, Xiao, and Lynch, 1993). These results establish that reversal is synapse specific (neighboring inputs were unaffected), occurs with small fiber populations, and involves a commonplace rhythm. Thirty sequential pulses are not likely to occur in vivo, but there is no reason to assume that fewer pulses would not be at least partially effective (30 was the lowest number used in the Larson, Xiao, and Lynch studies). It is also the case that, anec-dotally at least, reversal is more robust in vivo than in vitro. In all, conditions needed for reversal may occur routinely during behavior.

The vulnerability of memory to disruption during consolidation is not usually thought of with regard to its psychological utility. But the demonstration that near-normal physiologies can eliminate newly induced syn-aptic changes is certainly suggestive of a kind of "active forgetting." Something like this could be useful in focusing memory on essential features by erasing over trials the aspects of information that are not reliably detected in successive samplings. Forming a composite memory

FIGURE 10.7 Naturally occurring patterns of afferent activity ginning 30 seconds later; the other population (Pathway 1) did block consolidation. The illustrated experiments showed that not. This procedure was repeated five times at 2-min intervals.

the theta rhythm erases LTP as well as inducing it. Theta burst As shown in the bottom panel, robust and stable LTP devel-

pairs were delivered simultaneously to two small group of af- oped in Pathway 1, while Pathway 2 was barely changed.

ferents that converged on a population of target neurons (top (Modified from Larson, Xiao, and Lynch, 1993.) panel). One population (Pathway 2) received theta pulses be-

FIGURE 10.7 Naturally occurring patterns of afferent activity ginning 30 seconds later; the other population (Pathway 1) did block consolidation. The illustrated experiments showed that not. This procedure was repeated five times at 2-min intervals.

the theta rhythm erases LTP as well as inducing it. Theta burst As shown in the bottom panel, robust and stable LTP devel-

pairs were delivered simultaneously to two small group of af- oped in Pathway 1, while Pathway 2 was barely changed.

ferents that converged on a population of target neurons (top (Modified from Larson, Xiao, and Lynch, 1993.) panel). One population (Pathway 2) received theta pulses be-

of a building from different viewpoints could be an example. In any event, the idea that memories are being continuously made and discarded using two versions of the same rhythm (theta bursts, theta pulses) has consequences for network theories, the design of memory-enhancing drugs, and learning theory.

Active forgetting provides a first example in which LTP research suggests something about memory that has not been prominent in psychological studies or theories (the second of the upward links noted in the introductory paragraphs). Moreover, the idea arrives with details that would not have been found with behavioral analyses alone but that nonetheless have behavioral connotations. Theta pulses are an example: these are not detectable as such in behavior but are probably closely associated with a subset of psychological states (Vertis and Kocsis, 1997). Information of this sort not only fleshes out the idea of active forgetting but also provides a correlational approach for testing if it exists; for example, does the presence or absence of the pulses correlate with what is learned (e.g., the steepness of generalization gradients for a complex cue). Pharmacological manipulation provides another approach with which to search for active forgetting, but this is better considered in the general context of LTP and memory enhancement (see the next section). Finally, it bears repeating that the uncovering of interesting phenomena in response to developments at a more microscopic level of analysis has historically proved to be a critical step in the acceptance of theories about mechanisms. In this sense, reversal and forgetting may contribute to the debate on LTP and memory.

Was this article helpful?

0 0

Post a comment