General Precautions Necessary in Animal Models of Cognition

Many obvious precautions are described to ensure that performances on learning tasks are not biased. For example, studies are best conducted on young adult male rats. Rats that are not adults have immature nervous systems and may not learn consistently. Rats that are too old have age-related impairments that compromise their learning performances. Female rats experience estrus every 5 days, and their learning behavior may be influenced as a function of their hormonal status.

Rats should be obtained from the same batch for the entire study, otherwise heterogeneity across batches may confuse results. While the use of an inbred strain may to some extent ensure uniformity, there is no assurance that ever, within an inbred strain rats will be uniform in their behavior on a particular task (Pradhan et al., 1990). Only naive rats should be selected for experiments; rats that have been used in an earlier experiment are likely to show biases in behavior. The rats must be uniformly treated in matters ranging from housing to handling and feeding. If rats belonging to different experimental groups are treated differently, differing performances may be attributable to such differences in treatment rather than to differences in learning. The rats should be housed and maintained in reasonable comfort. Rats that are isolated one per cage, or that are otherwise stressed may perform poorly as a function of such stresses.

Rats should be handled regularly so that their responses to a learning task are not biased by the stress of the handling during the experiment. Rats should be familiarized with the experimental apparatus prior to the actual experiment so that their performances are not biased by exploratory behavior. The experiment must be conducted in an environment that is relatively sound proof and free from other distractions. The researcher must be seated such that his or her presence does not distract the rat. External stimuli, including lighting, should not cue the rat. Lighting in particular should be kept constant all through the study, because rats are very light sensitive, and become less or more motile with more and less environmental brightness, respectively. The apparatus must be cleaned after every rat has completed its task, otherwise the scent markings of the rat will bias the performances of future rats exposed to the apparatus. Learning assessments must be conducted at the same time of day lest circadian rhythm variations bias results. These and other precautions are well described in most textbooks on laboratory procedures (Bures et al., 1976; Joseph and Waddington, 1986; Van Ree and de Wied, 1988).

One further precaution deserves special mention. In animal research, rats are frequently assigned to one of several different groups. It is generally not feasible to complete an entire experiment in a single day; therefore, for convenience researchers sometimes execute their study by testing one group at a time. The fallacy of this procedure is that it permits the entry of sampling, handling, environmental, and other biases into the study. A more appropriate way of conducting the experiment is to ensure that each group is proportionately represented in each session of work.

ECS and Animal Models of Cognition

Models of ECS-Induced Amnesia

The literature on ECS and learning in animal models has been reviewed by Krueger et al. (1992) and Fochtmann (1994). This section will therefore provide only a brief summary. Retrograde amnesia associated with ECS has been studied most commonly using the passive avoidance paradigm (e.g., Alpern and McGaugh, 1968). Conditioned taste aversion has also been employed as a model (e.g., Shaw, 1986). Active avoidance, appetitive or aversive water reinforcement, bar pressing, conditioned emotional responses, T- and Y-maze learning, brightness discrimination, and hunger-fear conflict responses are some of the other paradigms that have been used to assess retrograde amnesia after single or repeated ECS (Fochtmann, 1994).

Anterograde amnesia with single or multiple ECS has been less extensively studied. Again, the most common method employed has been the passive avoidance paradigm (e.g., Gardner et al., 1972). Several other models of learning have also been described. Not all have succeeded in eliciting amnesia (Fochtmann, 1994). A general observation has been that the ability of ECS to prevent an association from occurring initially is more pronounced than its ability to disrupt an already formed association (Kral and Beggerly, 1973).

The method of ECS administration has been shown to affect the degree of memory impairment. Corneal electrode placement is associated with greater amnestic effects than transauricular electrode placement (Dorfman and Jarvik, 1968). Brief-pulse ECS is associated with less severe memory impairment as compared with sine wave stimuli (Docter, 1957). Altering the convulsion with the use of either anesthesia or nonconvulsive stimulation has variable effects on ECS-induced memory deficits. Increasing the number, frequency, intensity, or duration of ECS, or the proximity of the ECS to the time of training or testing, is associated with a greater degree of memory impairment (Fochtmann, 1994). These issues need to be kept in mind when choosing a model.

Practical Issues

Genetic differences influence task learning, and both good- and poor-learning strains have been discussed (Roullet and Lassalle, 1995; Van Buskirk and McGaugh, 1973). For example, C57BL/61bg mice are good learners in conditional spatial alternation tasks, while DBA/21bg mice are poor learners and require at least twice the number of training trials (Paylor, 1993). When an inbred strain of rats is unavailable, there is variation in learning behavior across batches of rats as well. Thus, previous experiences and textbook descriptions of animal models of cognition notwithstanding, prior to each experiment each laboratory may need to restandardize the model of learning and ECS-induced amnesia, readjusting variables ranging from the extent of pre-ECS training and the magnitude of aversive shocks to the strength, number, and frequency of ECS stimuli.