Harold C. Nielson

Evidence that electroconvulsive shock alters memory retrieval rather than memory consolidation

Abstract Four experiments reported here suggest that most experiments claiming that electroconvulsive shock (ECS) disrupts memory consolidation are confounded because of differences in activity levels and brain excitability states that exist between the experimental and control animals when they are tested for retention. Results of the first experiment showed that ECS increased open field activity levels. The effect of ECS upon brain excitability levels was determined in the second experiment by measuring the intensity of an electrical stimulus, delivered to a subcortical area, that would elicit a conditioned response. A transient decrease lasting 4 days was produced. Recovery of a learned response following ECS administration was investigated in the third experiment; the response was the avoidance of stepping off a platform. There was recovery when ECS induced increases in activity levels were controlled, and the recovery followed the same time course as the changes in brain excitability. A fourth experiment demonstrated that when activity levels and brain excitability states were equalized, ECS did not produce even a transient disruption of the avoidance response. These experiments show that ECS does not disrupt memory fixation which is dependent upon a neural reverberation process for consolidation, but suggest that memory retrieval may depend upon brain excitability states. The hypothesis is offered here that the neurological aspect of learning may involve changes in levels of brain excitability as reflected in the thresholds of functional neural systems, that retention implies a maintenance or reconstruction of these modifications of brain excitability, and that failure of retention occurs whenever brain excitability is modified away from that established by the training procedure.

Behavioral assessment of sodium arsanilate induced vestibular dysfunction in rats

Vestibular dysfunction was induced in Long-Evans and Wistar rats—previously thought to be refractory to the effects of ototoxic drugs—by subcutaneous or intratympanic injections of sodium arsanilate (atoxyl). Three measures of swimming behavior were taken to assess the degree of vestibular dysfunction. These measures were: escape latencies, errors, and underwater swimming times. Subcutaneous injections of sodium arsanilate that resulted in no behavioral impairment of previously trained rats resulted in less efficient learning of the maze by naive rats. Visual compensation for vestibular impairment was detected by altering or removing the visual cues the rats used in their compensation. A strain difference in sensitivity to subcutaneous injections of sodium arsanilate was found between the Wistar and Long-Evans rats, with the Wistar rats being much more sensitive.

Effects of electroconvulsive shock and prior stress on brain amine levels

Abstract An effort was made to determine whether the changes in activity levels of rats that received electroconvulsive shock (ECS) after either cold stress or foot shock could be related to changes in serotonin, norepinephrine, or dopamine levels in various parts of the brain. The brain regions sampled were the frontal pole, the caudate nucleus, and the remainder. There were some changes in the levels of these amines as a consequence of ECS treatment but the more significant changes were related to the type of stress the rats had been subjected to before ECS treatment. No clear relationships between changes in the levels of these amines in the brain and the reported changes in activity levels following ECS could be identified.

Cue use in state-dependent learning

Four experiments were conducted to determine what cues rats used to escape from a water maze in the normal nondrugged state or while drugged with pentobarbital. The subjects were trained to escape from a water maze that either required or enabled them to use place or response cues. The results showed that drugged and nondrugged rats used different cues to escape from the maze; drugged rats learned response sequences, while nondrugged rats learned either the place of exit from the maze or response sequences. Although some rats selected and used different cues in different states, they showed impaired learning when they were required to use the same cue in both the normal and drugged states. Learning in the drugged state appeared to be different from learning in the normal state, in that drugged rats were restricted in their cue use. The implications of these findings for theories of state-dependent learning were discussed.

Neural thresholds and state-dependent learning

Abstract In this experiment we tested the hypothesis that state-dependent learning is mediated by changes in brain excitability levels as determined by changes in the intensity of a conditioned stimulus (CS) delivered to the central nervous system, sufficient to maintain a conditioned response (CR). Four cats were trained in the normal state; CR thresholds determined for dosages of sodium pentobarbital ranging from 0 to 15 mg/kg; and then retrained to give CRs while in a drug state for 15 mg/kg pentobarbital. Another group of four cats was trained in the reverse order. They were trained to give CR in the drug state; thresholds taken as pentobarbital decreased from 15 mg/kg; and retrained in the normal state. Results showed that shifts in brain excitability levels cannot explain the state-dependent phenomenon; that the effect of pentobarbital upon the conditioning thresholds was a function of the animals prior experience with the drug and not directly a function of its pharmacological action; and that the drug and nondrug states were distinct before training, but not after training in both states.

Impairment of fixed-interval responding during chronic alcohol drinking in rats

Rats housed either in activity wheels or in standard cages were placed on a restricted feeding schedule and trained in operant chambers to respond on an FI-40 schedule for sugar-water rewards. When body weights and FI responding were stable, half of the subjects in each housing condition received both water and a 10% beer-ethanol solution ad lib for the next 30 days while the remaining subjects continued to receive water as their sole fluid. Subjects offered alcohol, irrespective of housing condition, drank a mean of 9.9 g/kg/day during the 30 days and showed shifts in the distribution of their responses within the fixed intervals. Reliable changes also occurred in measures of body weight, water intake, and activity for subjects that consumed alcohol.

A specific arousal effect from barpressing on nonspecific arousal in rats

Running-wheel activity and barpressing on either an FR 10 or FI 30-sec schedule of sugar-water reward were recorded, respectively, as measures of nonspecific and specific arousal of rats. The running-wheel activity of all rats was consistent across days and across within-day segments. Also, higher rates of running-wheel activity were associated with higher rates of FR performance but not FI performance. These results showed that nonspecific arousal had a positive relationship with both nonincentive- and incentive-motivated behaviors, although the relationship with the incentive task appeared to depend on whether efficient task performance could be facilitated or not by the higher levels of arousal. A separate effect of specific arousal also was in evidence since higher rates of activity on the running-wheel measure, relative to the previous night, were observed to follow higher rates of FR performance. The findings were interpreted as evidence for separate systems of arousal.

State-dependent dissociation of food consumption and maze running in rats

For 45 days, rats were given 30 min to consume wet mash immediately after they regained their righting reflexes following an IP injection of 25 mg/kg of sodium pentobarbital. Then they were trained to traverse a maze for a food reward while drugged, while not drugged, or 2 h before receiving the drug. After 28 days, the maze training continued but drug conditions were switched. The running latencies showed that running for a food reward was state-dependent. Food consumption also showed dissociation when the drugged condition was discontinued either at the start of maze training or during the drug reversal. Food consumption was not changed when the drugged condition was continued or reinstituted. Additional rats that received equal amounts of drug, but were not fed while drugged, showed that the effects on food consumption were related to the drugged-feeding experience and not to drug experience per se. A possible reorganization of sensory-motor integration associated with drugged feeding was discussed.

Indirect Effects of Footshock Stress and Coping on Alcohol Consumption of Food-Deprived and Non-Deprived Rats

Rats that lived in activity wheels were either food deprived or not food deprived and either allowed control or no control over footshock in a conditioning chamber. Measures of consumption of alcohol across successive phases of the experiment where footshock was administered or discontinued showed no direct effect of the footshock experiences on alcohol consumption. Food and water intake, body weight, and running-wheel activity were measured also, and direct effects of footshock on these measures were observed. Alcohol consumption was related to variables that reflected changes in caloric requirement. The possibility was discussed that footshocks and coping stresses affected a general food consumption or metabolic factor and any effect of the footshock procedures on alcohol intake was related to these factors primarily and to footshock only indirectly.

EFFECTS OF ANTICONVULSANT DRUGS UPON PATTERNS OF SEIZURE DISCHARGE AND BRAIN THRESHOLDS: SOME IMPLICATIONS FOR MEMORY MECHANISMS

10 cats were trained to perform an avoidance leg flexion at a signal: electrical stimulation of the caudate nucleus or the centre median. Thresholds were found to be elevated by anticonvulsants. In 3 of these cats, seizure discharges were induced by stimulation of the hippocampus and recorded on an EEG. The data are interpreted as indicating alteration of nerve impulse pathways. It is argued that such alteration may underlie dissociation of learning produced by various drugs and states.