Biochemical studies of learning and memory in goldfish : Effects of pretrial and posttrial administrations of stimulant and depressant drugs




Journal Title

Journal ISSN

Volume Title



Four experiments were conducted with the purpose of exploring the effects of CNS-active drugs on the learning and memory of a shuttle-box avoidance response in goldfish. Picrotoxin and sodium pentobarbital in dosages that did not cause apparent changes in activity level were employed for intracranial injection of 178 goldfish at different times relative to training. In the first experiment, three groups of goldfish were administered 6 μl of picrotoxin (S), sodium pentobarbital (D) or saline (C), respectively, five min prior to being given 20 trials of avoidance training. With the application of multiple nonparametric tests (MNT) to control the significance level experiment-wise, it was found that the picrotoxin group made significantly more, and the pentobarbital group significantly less avoidance responses than the saline group (a = 0.05). The modification of memory storage processes by these two drugs was implied. The additional test of activity showed that the drugs at the dosages used in the present study did not change the motor ability of goldfish. However, there could still be some side-effects of the drugs that contributed to the results, such as a change in sensitivity or/and attention. In order to evaluate the consolidation hypothesis in interpreting the drug effects in Exp I, 30 goldfish in Exp II were injected with the same doses immediately after each of three training sessions of 10 trials in the shuttlebox. They were tested by means of a relearning paradigm 48 hours after the previous session. Similar effects of the drugs on learning resulted from the post-trial injections, confirming the memory consolidation hypothesis. As the MNT showed, the facilitative effects of picrotoxin were evident in the second session after a single post-trial injection and the impairing effects of pentobarbital reached significance in the third session. No difference was found in the avoidance behaviors between the saline group and the picrotoxin group in the third session. Nevertheless, when the number of escapes was taken into account, the latter still performed better than the former. Picrotoxin, as a CNS stimulant, facilitates memory consolidation, while pentobarbital, as a CNS depressant, inhibits the consolidation process. Exp III was a study of the time-course of the effects of post-trial injections of these two drugs at the dosages chosen. As the first part showed, the results from immediate injection were consistent with those found in Exp II. A one-hour delayed injection was as effective as the immediate injection in enhancing learning in the case of picrotoxin [S(1)], but was not influential in the case of pentobarbital [D(1)]. Based on this result from part I, in an extended Part 2, two additional groups were trained. In one, picrotoxin was injected 4 hrs after the training session [S(4)], while in the other, pentobarbital injection was delayed 30 min [D(1/2)]. The former group avoided slightly better while the latter was slightly worse than the control group. When all the results in Exp III were put together and subjected to the statistical analysis with MNT no comparison reached significance in session 2 and only the difference between S(l) and D(0) yielded a significant result. Since S(l) and S(0), and D(l) and D(1/2) were not distinguishable from each other respectively, they were combined as two groups. MNT applied to the resulting five groups indicated that the group injected immediately after training with pentobarbital was significantly inferior to all of the other groups in session 3. The recording of cumulative latency showed that S(l), starting with the highest latency and more misses, learned to respond more readily with shorter latencies in the second session than most of the other groups. From the results of Exp III, it was inferred that the 'threshold time' in which a drug was effective in modifying the memory consolidation process of the shuttle-box avoidance response was longer for 1.5 mg/kg picrotoxin than that for 30 mg/kg pentobarbital. Although the immediate injection of pentobarbital almost totally blocked the retention, the delayed injection was ineffective. Exp IV was a test of the drug effects on the extinction process. However, due to the factor of using Ss discarded from Exp III, the goldfish did not extinguish their avoidance behavior to the light. Nevertheless, it lends more evidence to exclude the possibility of cumulative drug effects. The possibility of proactive effects of drugs on learning was examined and excluded. No consistent relation could be drawn between general activity level and avoidance behavior. The results of this study confirmed and extended the previous CNS-active drug studies in rodent animals. Goldfish once again proved to be ideal Ss for biochemical studies of learning and memory.



Learning in animals, Memory transfer, Psychopharmacology, Fishes--Training